Rapid Ampicillin Susceptibility Testing for Haemophilus influenzae ROGER M. BARKIN, M.D., JAMES K. TODD, M.D., AND MARTHA H. ROE, M.T.(ASCP)
RECENT REPORTS of strains of Haemophilus influenzae resistant to ampicillin make rapid, accurate antibiotic susceptibility testing imperative in the management of systemic Haemophilus disease.711 The existence of several acceptable antibiotic regimens other than ampicillin1,5'8 provides the clinician with therapeutic alternatives when dealing with Haemophilus infections, but rapid susceptibility results could permit more optimal drug selection and the early elimination of unnecessary second drugs. Received November 10, 1975; received revised manuscript February 25, 1976; accepted for publication February 25, 1976. Presented in part at the Society for Pediatric Research Annual Meeting, Denver, Colorado, April 18, 1975. Address reprint requests to Dr. Todd: Director of Infectious Disease, The Children's Hospital of Denver, 1056 E. 19th Ave., Denver, Colorado 80218.
From the Departments of Pathology and Pediatrics, The Children's Hospital of Denver, and the Department of Pediatrics, University of Colorado Medical Center, Denver, Colorado
Several effective means of determining ampicillin susceptibility have been described, but all require isolation of the organisms prior to testing.2,3,9 The modified Kirby-Bauer technic has been used extensively but is poorly standardized for Haemophilus influenzae and may indicate false resistance.2 Recently reported simplified microtiter determinations of minimal inhibitory concentrations (MIC) again require the prior isolation of the organism, as well as considerable laboratory expertise in interpretation.3 A rapid beta-lactamase detection method has been described9 but indicated false ampicillin susceptibility in 10% of the resistant Haemophilus strains such that the authors suggested the need for confirmation by other technics. To provide rapid ampicillin susceptibility testing, an agar-dilution method permitting reliable susceptibility determinations on clinical specimens 6-18 hours after receipt of the specimens in the laboratory was developed. Materials and Methods Organisms Beginning in June 1974, through December 1975 (18 months), all clinical specimens ultimately growing Haemophilus influenzae had ampicillin susceptibility tested by direct application of the clinical specimen to the agar biplates as well as by the standard disk diffusion method. Twenty-seven test strains of Haemophilus influenzae, type B, were further studied in detail. These consisted of 21 of the above-mentioned susceptible organisms, 11 from cerebrospinal fluid, eight from blood cultures,
100
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Barkin, Roger M., Todd, James K., and Roe, Martha H.: Rapid ampicillin susceptibility testing for Haemophilus influenzae. Am J Clin Pathol 67: 100-103, 1977. Recent isolations of strains of Haemophilus influenzae resistant to ampicillin necessitate the development of a rapid, dependable, reproducible method of determining their antibiotic susceptibility. An agar-dilution method permitting susceptibility determinations on clinical specimens within 6-18 hours of specimen collection was designed. Chocolate agar biplates were made with one side having no additive and the other containing 2 jug/ml ampicillin. Seventy clinical specimens (cerebrospinal fluid, joint fluid, car fluid, pleural fluid, blood culture broth) were streaked directly onto both sides of the plates when received in the laboratory and incubated at 35-37 C in 10% CO-2. Reliable, readable results were usually available within 6-18 hours of receipt of the specimen and correlated completely with minimum inhibitory concentrations (MIC) determined by the agar-dilution plate method, although standard disk susceptibilities occasionally indicated false resistance. Susceptible strains (MIC < 2 /xg/ml) grew on the antibiotic-free side of the biplate only. The rapid determination of ampicillin susceptibility allows optimal antibiotic selection for the treatment of Haemophilus influenzae infections with early discontinuation of potentially toxic supplementary drugs. (Key words: Haemophilus influenzae; Ampicillin resistance; Antibiotic susceptibility testing.)
Vol. 67 • No. I
RAPID AMPICILLIN
101
Fie. I. Agar biplate streaked on both sides with an ampicillin-susceptible strain of Haemophilus influenzae. Growth occurred only on the side not containing ampicillin (2 Mg/ml).
Media The agar-dilution biplates and the plates for agar plate dilution and disk diffusion testing were made of Mueller-Hinton agar base with 5% chocolatized rabbit blood and 1% IsoVitaleX (BBL). Ampicillin determinations were made using Medium 5 (Difco). Agar Dilution Biplates Agar dilution biplates were prepared by placing the standard chocolate base agar on half of the plate and adding an appropriate concentration of ampicillin to the molten (45 C) agar prior to pouring the other side, producing a final concentration of 2 /u,g/ml ampicillin. Known resistant and susceptible strains of Haemophilus influenzae were tested with each new batch of plates to assure adequate quality control. Plates were stored in sealed bags for as long as ten days at 4 C. The biplates were incubated aerobicaily at 35-37 C in 10% C0 2 and inspected every 6 hours for growth. In addition, adjusted inocula of the test strains ranging from 104 to
108 colony-forming units/ml (CFU/ml) were tested for growth on the biplates. Inocula Clinical specimens were plated directly on the biplates. MIC and disk susceptibility determinations used inocula standardized photometrically in MuellerHinton broth, following overnight growth on chocolate Mueller-Hinton plates with 1% IsoVitaleX. The inocula were adjusted to approximately 108 CFU/ml for disk diffusion testing and 10", 10s, and 108 CFU/ml for agardilution MIC determinations. The same concentrations were tested using the agar-dilution biplates, in addition to the direct plating of clinical specimens. MIC Determination Agar-dilution plates were prepared by the addition of appropriate ampicillin concentrations to molten (45 C) chocolate agar. Concentrations ranged from 0.2 to 32 /xg/ml. Using a Steers replicator, 0.01 ml of each strain of Haemophilus influenzae in concentrations of 104, 106, and 108 CFU/ml was applied in triplicate as a single drop to the agar surface of an ampicillin-containing plate of each concentration. A positive control plate was quantitatively inoculated to confirm inoculum size.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
and one each from joint and pleural fluids. Six resistant organisms were investigated (five obtained from the Center for Disease Control, CDC), four from cerebrospinal fluid and one each from ear and sinus aspirates.
BARKIN, TODD AND ROE
102
inhibition were measured. Ampicillin concentrations of the wedges were then calculated. Levels were determined on the biplates refrigerated over a 14-day period, as well as after 24 hours of incubation at 35-37 C.
8.0 i 6.0 ^ E
4.0 Ampicillin Agar Dilution Biplate (2 ugm/ml)
-? 2.0 4
z
Disk Diffusion Testing2
= 0.8 H I
A.J.C.P. • January 1977
0.4-
I I •• t
0.2-
I
I
1
1
l
I
l
I
l
l
l
l
8
10
12
14
16
18
20
22
24
26
28
30
Disc
Zone (mm)
FIG. 2. Ampicillin-resistant and ampicillin-susceptible strains of Haemophilus influenzae—disk zone size compared with MIC.
A cotton swab moistened with a standardized suspension (108 CFU/ml) of Haemophilus influenzae organisms was used to produce a uniform lawn on chocolate agar plates. A 10-/xg ampicillin disk (Pfizer) was applied to the plate, and the plates were incubated for 24 hours at 35-37 C in 10% C0 2 . The diameter of the zone of inhibition was measured. Results represent the average of two separate determinations.
1
1.90
>^»—
1, 1.80 3.
\
\
= 1.70
t t
* \
O
\ \\
2 1.60C
S 1.50 t
\
\
o
1.40
\ \
\ \\
\
\ \
b
—
h
>a. 1.30
Incubated for 24 hours (37°C)
•1 1.201
2
3
4
5
6
7
8
9
10
11
12
13
14
Days of Storage
FIG. 3. Stability of ampicillin in agar biplates stored at 4 C and incubated at 37 C.
The plates were incubated for 24 hours at 35-37 C in 10% C0 2 . The MIC was considered to be the lowest concentration of ampicillin significantly inhibiting growth of the inoculum compared with the control. Results represent the average of triplicate MIC determinations. Stability of Ampicillin in Biplates The ampicillin concentrations6 in the agar-dilution biplates over a period of 14 days of storage at 4 C in sealed bags were determined. Wedges of ampicillinchocolate agar were weighed, mashed, and added to known volumes of Mueller-Hinton broth. After centrifugation, an aliquot of the supernatant was placed on a blank paper disk (BBL). Known concentrations of ampicillin diluted in Mueller-Hinton broth were placed on similar paper disks. The disks were then placed on Medium 5 (Difco) to which 0.01 ml of Bacillus subtilis spore suspension (108 spores/ml) was added. After 4 hours of incubation at 35-37 C, the zones of growth
The agar-dilution biplates clearly distinguished the ampicillin-resistant from the susceptible strains of Haemophilus influenzae, type B, when the 70 clinical specimens were inoculated directly on the ampicillin biplates. The only resistant strain grew on both sides of the biplate, while the 69 susceptible strains grew only on the side not containing 2 /ug/ml ampicillin (Fig. 1). Six of the susceptible strains were initially thought to be resistant to ampicillin on the basis of results of the standard disk diffusion method. Within 18 hours of inoculation of clinical specimens, the organisms were clearly shown to be susceptible or resistant and appropriate clinical changes in antibiotic therapy could be made. Inocula of the 27 test strains in the range of 104 — 10s CFU/ml could clearly be defined as susceptible or resistant to ampicillin on the biplates. Less than 103 CFU/ml in the inocula (ten organisms per plate) produced too few colonies to make that distinction clearly, Table 1. Agar Dilution MIC's of Haemophilus influenzae Test Strains MIC (jug/ml)
Microtiter(CDC) Agar dilution* (0.3-1.0 x 104 CFU/ml) Agar dilution* (0.3-1.0 x 106 CFU/ml) Agar dilution* (0.3-1.0 x 108 CFU/ml)
Susceptible Strains
Resistant Strains
64
N.D. N.D. N.D. — — 3 16
5
—
14
7
—
3
12
9
—
— — 1
* End point = significant inhibition of growth.
— 4 2 — — — 3
1
4
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Results
2.00-
103
RAPID AMPICILLIN
Vol. 67 • No. I
8
Discussion The agar-dilution biplates have proven to be a reliable, rapid, and reproducible method of determining ampicillin susceptibility of Haemophilus influenzae. Clinical specimens in which gram-negative pleomorphic rods are found on Gram stain can be plated directly without isolation or subculturing, providing susceptibility data within 6-18 hours of inoculation. Using the early subculture method, 10 blood cultures from patients suspected to have Haemophilus influenzae bacteremia can be subcultured onto both sides of the ampicillin-agar biplates and ampicillin susceptibility results can be rapidly determined in the initial sub-
culture. Plates remain stable for 11 days when stored at 4 C in sealed bags. The level of ampicillin in the biplates may necessitate some quantitation of the inoculum, but the inoculum effect is apparently greater in resistant than in susceptible organisms, such that resistant strains will not be missed. Recent concern over the appearance of ampicillinresistant strains of Haemophilus influenzae, type B, has resulted in the recommendation that additional supplementary antibiotic therapies be used until accurate antibiotic susceptibility determinations can be made. 4 Since the alternative modes of therapy (chloramphenicol, 8 tetracycline, 5 streptomycin 1 ) may be associated with significant risks and side effects in children, it is reasonable to make alterations in treatment as soon as possible, utilizing accurate antibiotic susceptibility data. The ampicillin-agar biplate provides accurate susceptibility determinations, allowing appropriate alterations of antibiotic therapy within 6-18 hours of initiation of treatment, in contrast to the disk and MIC methods, which give results in a minimum of 48-72 hours, the former occasionally yielding false-resistant results. References 1. Appelbaum E, Nelson J: Streptomycin in the treatment of influenzae meningitis. JAMA 143:715. 1952 2. Center for Disease Control: Antimicrobial sensitivity tests for Haemophilus influenzae. Morbidity and Mortality Weekly Report 23:11, 1974 3. Jorgensen JH, Jones PM: Simplified medium for susceptibility testing of Haemophilus influenzae. Antimicrob Ag Chemother 7:186, 1975 4. Katz SL: Ampicillin-resistant Haemophilus influenzae type B: A status report. Pediatrics 55:6. 1975 5. Nelson KE, Levin S, Spies HW: Treatment of Haemophilus influenzae meningitis; a comparison of chloramphenicol and tetracycline. J Infect Dis 125:459, 1972 6. Sabath LD: A simple, rapid microassay for nephrotoxic antibiotics. Scope Monograph (S-2061). Upjohn Co.. Kalamazoo, 1972 7. Schiffer MS, Schneerson R. MacLowry J, et al: Clinical, bacteriological and immunological characterisation of ampicillin-resistant Haemophilus influenzae type B. Lancet 2:257, 1974 8. Shackleford PG, Bobinski JE, Feigin RD. et al: Therapy of Haemophilus influenzae meningitis reconsidered. N Engl J Med 287:634. 1972. 9. Thornsberry C, Kirven LA: Ampicillin resistance to Haemophilus influenzae as determined by a rapid test for betalactamase production. Antimicrob Ag Chemother 6:653, 1974 10. Todd JK, Roe MH: Rapid detection of bacteremia by an early subculture technique. Am J Clin Pathol 64:694-699, 1975 11. Tomeh MO, Starr SE, McGowan TE, et al: Ampicillinresistant Haemophilus influenzae type B infection. JAMA 229:295, 1974
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
and inocula equal to or greater than 10 CFU/ml occasionally produced some growth on the ampicillin side of the biplate, although a quantitative difference in growth could still be identified. The concentrations of Haemophilus influenzae in clinical specimens were always less than I08 CFU/ml. MIC determinations on the test strain organisms by agar dilution were performed using inoculum concentrations of I04, I06, and 108 CFU/ml (Table 1). A definite inoculum effect was seen especially at the highest organism concentration (108 CFU/ml), but the distinction between susceptible and resistant Haemophilus influenzae was clearly maintained. MIC determinations documented that susceptible organisms were inhibited by concentrations well below 2 fxglm\ (less than or equal to 0.2-0.4 /xg/ml) and resistant ones had MIC's well above this level (greater than or equal to 4.0 ^ig/ml). Disk diffusion testing of the test organisms by the modified Kirby-Bauer method produced growth inhibition zones ranging from 20 to 30 mm for the susceptible organisms and from 13 to 18 mm for the resistant organisms after 24 hours of incubation. Figure 2 shows the relationship between resistant and susceptible test strains of Haemophilus influenzae as determined by conventional disk diffusion methods and agar-dilution MIC determinations. Ampicillin levels were maintained for 11 days in the biplates, with some diminution in levels present beginning on the twelfth day of storage. A 24-30% decrease in ampicillin level was seen following 24 hours of incubation at 35-37 C in 10% C 0 2 after any period of storage (Fig. 3).
RECENT REPORTS of strains of Haemophilus influenzae resistant to ampicillin make rapid, accurate antibiotic susceptibility testing imperative in the management of systemic Haemophilus disease.711 The existence of several acceptable antibiotic regimens other than ampicillin1,5'8 provides the clinician with therapeutic alternatives when dealing with Haemophilus infections, but rapid susceptibility results could permit more optimal drug selection and the early elimination of unnecessary second drugs. Received November 10, 1975; received revised manuscript February 25, 1976; accepted for publication February 25, 1976. Presented in part at the Society for Pediatric Research Annual Meeting, Denver, Colorado, April 18, 1975. Address reprint requests to Dr. Todd: Director of Infectious Disease, The Children's Hospital of Denver, 1056 E. 19th Ave., Denver, Colorado 80218.
From the Departments of Pathology and Pediatrics, The Children's Hospital of Denver, and the Department of Pediatrics, University of Colorado Medical Center, Denver, Colorado
Several effective means of determining ampicillin susceptibility have been described, but all require isolation of the organisms prior to testing.2,3,9 The modified Kirby-Bauer technic has been used extensively but is poorly standardized for Haemophilus influenzae and may indicate false resistance.2 Recently reported simplified microtiter determinations of minimal inhibitory concentrations (MIC) again require the prior isolation of the organism, as well as considerable laboratory expertise in interpretation.3 A rapid beta-lactamase detection method has been described9 but indicated false ampicillin susceptibility in 10% of the resistant Haemophilus strains such that the authors suggested the need for confirmation by other technics. To provide rapid ampicillin susceptibility testing, an agar-dilution method permitting reliable susceptibility determinations on clinical specimens 6-18 hours after receipt of the specimens in the laboratory was developed. Materials and Methods Organisms Beginning in June 1974, through December 1975 (18 months), all clinical specimens ultimately growing Haemophilus influenzae had ampicillin susceptibility tested by direct application of the clinical specimen to the agar biplates as well as by the standard disk diffusion method. Twenty-seven test strains of Haemophilus influenzae, type B, were further studied in detail. These consisted of 21 of the above-mentioned susceptible organisms, 11 from cerebrospinal fluid, eight from blood cultures,
100
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Barkin, Roger M., Todd, James K., and Roe, Martha H.: Rapid ampicillin susceptibility testing for Haemophilus influenzae. Am J Clin Pathol 67: 100-103, 1977. Recent isolations of strains of Haemophilus influenzae resistant to ampicillin necessitate the development of a rapid, dependable, reproducible method of determining their antibiotic susceptibility. An agar-dilution method permitting susceptibility determinations on clinical specimens within 6-18 hours of specimen collection was designed. Chocolate agar biplates were made with one side having no additive and the other containing 2 jug/ml ampicillin. Seventy clinical specimens (cerebrospinal fluid, joint fluid, car fluid, pleural fluid, blood culture broth) were streaked directly onto both sides of the plates when received in the laboratory and incubated at 35-37 C in 10% CO-2. Reliable, readable results were usually available within 6-18 hours of receipt of the specimen and correlated completely with minimum inhibitory concentrations (MIC) determined by the agar-dilution plate method, although standard disk susceptibilities occasionally indicated false resistance. Susceptible strains (MIC < 2 /xg/ml) grew on the antibiotic-free side of the biplate only. The rapid determination of ampicillin susceptibility allows optimal antibiotic selection for the treatment of Haemophilus influenzae infections with early discontinuation of potentially toxic supplementary drugs. (Key words: Haemophilus influenzae; Ampicillin resistance; Antibiotic susceptibility testing.)
Vol. 67 • No. I
RAPID AMPICILLIN
101
Fie. I. Agar biplate streaked on both sides with an ampicillin-susceptible strain of Haemophilus influenzae. Growth occurred only on the side not containing ampicillin (2 Mg/ml).
Media The agar-dilution biplates and the plates for agar plate dilution and disk diffusion testing were made of Mueller-Hinton agar base with 5% chocolatized rabbit blood and 1% IsoVitaleX (BBL). Ampicillin determinations were made using Medium 5 (Difco). Agar Dilution Biplates Agar dilution biplates were prepared by placing the standard chocolate base agar on half of the plate and adding an appropriate concentration of ampicillin to the molten (45 C) agar prior to pouring the other side, producing a final concentration of 2 /u,g/ml ampicillin. Known resistant and susceptible strains of Haemophilus influenzae were tested with each new batch of plates to assure adequate quality control. Plates were stored in sealed bags for as long as ten days at 4 C. The biplates were incubated aerobicaily at 35-37 C in 10% C0 2 and inspected every 6 hours for growth. In addition, adjusted inocula of the test strains ranging from 104 to
108 colony-forming units/ml (CFU/ml) were tested for growth on the biplates. Inocula Clinical specimens were plated directly on the biplates. MIC and disk susceptibility determinations used inocula standardized photometrically in MuellerHinton broth, following overnight growth on chocolate Mueller-Hinton plates with 1% IsoVitaleX. The inocula were adjusted to approximately 108 CFU/ml for disk diffusion testing and 10", 10s, and 108 CFU/ml for agardilution MIC determinations. The same concentrations were tested using the agar-dilution biplates, in addition to the direct plating of clinical specimens. MIC Determination Agar-dilution plates were prepared by the addition of appropriate ampicillin concentrations to molten (45 C) chocolate agar. Concentrations ranged from 0.2 to 32 /xg/ml. Using a Steers replicator, 0.01 ml of each strain of Haemophilus influenzae in concentrations of 104, 106, and 108 CFU/ml was applied in triplicate as a single drop to the agar surface of an ampicillin-containing plate of each concentration. A positive control plate was quantitatively inoculated to confirm inoculum size.
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
and one each from joint and pleural fluids. Six resistant organisms were investigated (five obtained from the Center for Disease Control, CDC), four from cerebrospinal fluid and one each from ear and sinus aspirates.
BARKIN, TODD AND ROE
102
inhibition were measured. Ampicillin concentrations of the wedges were then calculated. Levels were determined on the biplates refrigerated over a 14-day period, as well as after 24 hours of incubation at 35-37 C.
8.0 i 6.0 ^ E
4.0 Ampicillin Agar Dilution Biplate (2 ugm/ml)
-? 2.0 4
z
Disk Diffusion Testing2
= 0.8 H I
A.J.C.P. • January 1977
0.4-
I I •• t
0.2-
I
I
1
1
l
I
l
I
l
l
l
l
8
10
12
14
16
18
20
22
24
26
28
30
Disc
Zone (mm)
FIG. 2. Ampicillin-resistant and ampicillin-susceptible strains of Haemophilus influenzae—disk zone size compared with MIC.
A cotton swab moistened with a standardized suspension (108 CFU/ml) of Haemophilus influenzae organisms was used to produce a uniform lawn on chocolate agar plates. A 10-/xg ampicillin disk (Pfizer) was applied to the plate, and the plates were incubated for 24 hours at 35-37 C in 10% C0 2 . The diameter of the zone of inhibition was measured. Results represent the average of two separate determinations.
1
1.90
>^»—
1, 1.80 3.
\
\
= 1.70
t t
* \
O
\ \\
2 1.60C
S 1.50 t
\
\
o
1.40
\ \
\ \\
\
\ \
b
—
h
>a. 1.30
Incubated for 24 hours (37°C)
•1 1.201
2
3
4
5
6
7
8
9
10
11
12
13
14
Days of Storage
FIG. 3. Stability of ampicillin in agar biplates stored at 4 C and incubated at 37 C.
The plates were incubated for 24 hours at 35-37 C in 10% C0 2 . The MIC was considered to be the lowest concentration of ampicillin significantly inhibiting growth of the inoculum compared with the control. Results represent the average of triplicate MIC determinations. Stability of Ampicillin in Biplates The ampicillin concentrations6 in the agar-dilution biplates over a period of 14 days of storage at 4 C in sealed bags were determined. Wedges of ampicillinchocolate agar were weighed, mashed, and added to known volumes of Mueller-Hinton broth. After centrifugation, an aliquot of the supernatant was placed on a blank paper disk (BBL). Known concentrations of ampicillin diluted in Mueller-Hinton broth were placed on similar paper disks. The disks were then placed on Medium 5 (Difco) to which 0.01 ml of Bacillus subtilis spore suspension (108 spores/ml) was added. After 4 hours of incubation at 35-37 C, the zones of growth
The agar-dilution biplates clearly distinguished the ampicillin-resistant from the susceptible strains of Haemophilus influenzae, type B, when the 70 clinical specimens were inoculated directly on the ampicillin biplates. The only resistant strain grew on both sides of the biplate, while the 69 susceptible strains grew only on the side not containing 2 /ug/ml ampicillin (Fig. 1). Six of the susceptible strains were initially thought to be resistant to ampicillin on the basis of results of the standard disk diffusion method. Within 18 hours of inoculation of clinical specimens, the organisms were clearly shown to be susceptible or resistant and appropriate clinical changes in antibiotic therapy could be made. Inocula of the 27 test strains in the range of 104 — 10s CFU/ml could clearly be defined as susceptible or resistant to ampicillin on the biplates. Less than 103 CFU/ml in the inocula (ten organisms per plate) produced too few colonies to make that distinction clearly, Table 1. Agar Dilution MIC's of Haemophilus influenzae Test Strains MIC (jug/ml)
Microtiter(CDC) Agar dilution* (0.3-1.0 x 104 CFU/ml) Agar dilution* (0.3-1.0 x 106 CFU/ml) Agar dilution* (0.3-1.0 x 108 CFU/ml)
Susceptible Strains
Resistant Strains
64
N.D. N.D. N.D. — — 3 16
5
—
14
7
—
3
12
9
—
— — 1
* End point = significant inhibition of growth.
— 4 2 — — — 3
1
4
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
Results
2.00-
103
RAPID AMPICILLIN
Vol. 67 • No. I
8
Discussion The agar-dilution biplates have proven to be a reliable, rapid, and reproducible method of determining ampicillin susceptibility of Haemophilus influenzae. Clinical specimens in which gram-negative pleomorphic rods are found on Gram stain can be plated directly without isolation or subculturing, providing susceptibility data within 6-18 hours of inoculation. Using the early subculture method, 10 blood cultures from patients suspected to have Haemophilus influenzae bacteremia can be subcultured onto both sides of the ampicillin-agar biplates and ampicillin susceptibility results can be rapidly determined in the initial sub-
culture. Plates remain stable for 11 days when stored at 4 C in sealed bags. The level of ampicillin in the biplates may necessitate some quantitation of the inoculum, but the inoculum effect is apparently greater in resistant than in susceptible organisms, such that resistant strains will not be missed. Recent concern over the appearance of ampicillinresistant strains of Haemophilus influenzae, type B, has resulted in the recommendation that additional supplementary antibiotic therapies be used until accurate antibiotic susceptibility determinations can be made. 4 Since the alternative modes of therapy (chloramphenicol, 8 tetracycline, 5 streptomycin 1 ) may be associated with significant risks and side effects in children, it is reasonable to make alterations in treatment as soon as possible, utilizing accurate antibiotic susceptibility data. The ampicillin-agar biplate provides accurate susceptibility determinations, allowing appropriate alterations of antibiotic therapy within 6-18 hours of initiation of treatment, in contrast to the disk and MIC methods, which give results in a minimum of 48-72 hours, the former occasionally yielding false-resistant results. References 1. Appelbaum E, Nelson J: Streptomycin in the treatment of influenzae meningitis. JAMA 143:715. 1952 2. Center for Disease Control: Antimicrobial sensitivity tests for Haemophilus influenzae. Morbidity and Mortality Weekly Report 23:11, 1974 3. Jorgensen JH, Jones PM: Simplified medium for susceptibility testing of Haemophilus influenzae. Antimicrob Ag Chemother 7:186, 1975 4. Katz SL: Ampicillin-resistant Haemophilus influenzae type B: A status report. Pediatrics 55:6. 1975 5. Nelson KE, Levin S, Spies HW: Treatment of Haemophilus influenzae meningitis; a comparison of chloramphenicol and tetracycline. J Infect Dis 125:459, 1972 6. Sabath LD: A simple, rapid microassay for nephrotoxic antibiotics. Scope Monograph (S-2061). Upjohn Co.. Kalamazoo, 1972 7. Schiffer MS, Schneerson R. MacLowry J, et al: Clinical, bacteriological and immunological characterisation of ampicillin-resistant Haemophilus influenzae type B. Lancet 2:257, 1974 8. Shackleford PG, Bobinski JE, Feigin RD. et al: Therapy of Haemophilus influenzae meningitis reconsidered. N Engl J Med 287:634. 1972. 9. Thornsberry C, Kirven LA: Ampicillin resistance to Haemophilus influenzae as determined by a rapid test for betalactamase production. Antimicrob Ag Chemother 6:653, 1974 10. Todd JK, Roe MH: Rapid detection of bacteremia by an early subculture technique. Am J Clin Pathol 64:694-699, 1975 11. Tomeh MO, Starr SE, McGowan TE, et al: Ampicillinresistant Haemophilus influenzae type B infection. JAMA 229:295, 1974
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
and inocula equal to or greater than 10 CFU/ml occasionally produced some growth on the ampicillin side of the biplate, although a quantitative difference in growth could still be identified. The concentrations of Haemophilus influenzae in clinical specimens were always less than I08 CFU/ml. MIC determinations on the test strain organisms by agar dilution were performed using inoculum concentrations of I04, I06, and 108 CFU/ml (Table 1). A definite inoculum effect was seen especially at the highest organism concentration (108 CFU/ml), but the distinction between susceptible and resistant Haemophilus influenzae was clearly maintained. MIC determinations documented that susceptible organisms were inhibited by concentrations well below 2 fxglm\ (less than or equal to 0.2-0.4 /xg/ml) and resistant ones had MIC's well above this level (greater than or equal to 4.0 ^ig/ml). Disk diffusion testing of the test organisms by the modified Kirby-Bauer method produced growth inhibition zones ranging from 20 to 30 mm for the susceptible organisms and from 13 to 18 mm for the resistant organisms after 24 hours of incubation. Figure 2 shows the relationship between resistant and susceptible test strains of Haemophilus influenzae as determined by conventional disk diffusion methods and agar-dilution MIC determinations. Ampicillin levels were maintained for 11 days in the biplates, with some diminution in levels present beginning on the twelfth day of storage. A 24-30% decrease in ampicillin level was seen following 24 hours of incubation at 35-37 C in 10% C 0 2 after any period of storage (Fig. 3).
