The antimicrobial activity in vitro of chlorhexidine, a mixture of isothiazolinones (‘Kathon’ CG) and cetyl trimethyl ammonium bromide (CTAB)

j’ournal

of Hospital

Infection

(1993) 23, 87-l 11

The antimicrobial activity in v&o of chlorhexidine, a mixture of isothiazolinones (‘Kathon’ CC) and cetyl trimethyl ammonium bromide (CTAB) G. Nicoletti,

V. Boghossian, F. Gurevitch, P. Morgenroth

Department of Applied Biology of Technology, 124 Latrobe

R. Borland

and

& Biotechnology, Royal Melbourne Institute St, Melbourne, 3000, Victoria, Australia

Accepted for publication

8 December 1992

Summary: Chlorhexidine, two 4% chlorhexidine antiseptic handwashes (‘Bioprep’ and ‘Hibiclens’), cetyl trimethyl ammonium bromide (CTAB) and isothiazolinones (‘Kathon’) were tested against Staphylococcus aureus, Micrococcus luteus, Escherichia coli, Serratia marcescens, Pseudomonas aeruginosa, Proteus vulgaris and Candida albicans. The activities measured were the minimum inhibitory concentration (MIC), minimum microbicidal concentration (MMC). rate of kill in water and broth. effect of organic soil. I the development of microbial resistance on continuous exposure and agent bioavailability in media and formulation. ‘Kathon’ was the most active microbistatic agent showing maximal activity at low concentration, least inactivation by organic soil and media components and the lowest level of development of bacterial resistance. It was synergistic with chlorhexidine against S. marcescens and P. aeruginosa. Media, formulation components and organic soil affected the performance of chlorhexidine and CTAB. Chlorhexidine was more broadly active than CTAB but showed a greater reduction in activity in the presence of soil and engendered a greater level of bacterial resistance. It was more rapidly bactericidal to P. aeruginosa and S. marcescens than to S. aureus. Stable resistance to chlorhexidine and CTAB was developed by P. aeruginosa and S. marcescens, the latter showing the higher level of resistance. Chlorhexidine-resistant strains were also resistant to CTAB. The antiseptic formulations were more rapidly bactericidal than chlorhexidine alone but were otherwise of comparable activity. Mixtures of disinfectants, in particular a combination of chlorhexidine and a preservative level of ‘Kathon’, were more active than single disinfectants. The importance of standardization of media and test conditions and the use of chemically defined media for accurate and reproducible in-vitro testing of disinfectant activity is emphasized. Disinfection kinetics, expressed as time-kill curves, log reduction factors or decimal reduction times were shown to be valuable in differentiating microbistatic from microbicidal activity, showing the effects of dilution and soil on activity and indicating possible different mechanisms of action. Keywords: Chlorhexidine; ‘Kathon’; cetyl disinfectant evaluation; disinfection kinetics; bioavailability.

Correspondence 01956701~93,'l~20087

to: MS G. Nicoletti. + 25 $0X ,,O:O

trimethyl bacterial

ammonium bromide; resistance; disinfectant

G. Nicoletti

et al.

Introduction

We report here on the antimicrobial activity of chlorhexidine, cetyl trimethyl ammonium bromide, isothiazolinones (‘Kathon’ CG) and two chlorhexidine antiseptic handwash formulations. We have investigated the level and rate of microbial activity, the effect of medium and formulation components and organic soil on activity and the development of microbial resistance to the agents. The disinfectants tested were used as controls in the evaluation of new agents being developed for use as disinfectants and antiseptics. Test microorganisms were selected for their significance in hospital infections and in contamination of antiseptic preparations. Chlorhexidine, a biguanide, is widely used in medical and veterinary applications. It is a stable, strongly cationic compound which binds readily to organic and inorganic anions. Its antimicrobial activity and mechanism of action have been widely investigated and reviewed.‘z2 It is rapidly microbicidal against a wide range of Gram-positive and Gram-negative bacteria, yeasts and moulds. Activity is reduced in the presence of organic soil and high concentrations of inorganic and organic anions and non-ionic surfactants. Chlorhexidine affects the cytoplasmic membrane, inducing leakage of ions solutes and inhibiting certain and cytoplasmic membrane-bound enzymes, such as adenosyl triphosphatase; also at high concentrations it precipitates proteins and nucleic acids. ‘Kathon’ CG (cosmetic grade) is a 3:l mixture of isothiazolinones (S-chloro-2-methyl-4-isothiazolin-3-one and 2-methyl-4-isothiazolin-3one) stabilized with MgCl, (9%) and Mg (NO,), (16oh).3-5 The mixture is a broad spectrum microbistatic agent effective in the range l-10 mg 1-l against Gram-positive and Gram-negative bacteria, fungi and algae. ‘Kathon’ is odourless and colourless, stable at pH d 8, effective over a wide pH range but more active at acidic pH, miscible with water, lower alcohols and glycols, not inactivated by anionic or cationic molecules but inactivated by high levels of thiol containing compounds like glutathione and sulphur-containing amino acids. h-7 The isothiazolinones are postulated to exert their inhibitory effects by reversibly interacting with thiol groups on vital cell molecules.7 Toxicological evaluation and environmental impact studies have been reported. 5,8-1o It is widely used as a preservative and several different formulations are commercially available. It is used at 3-l 5 ppm, in many cosmetic and toiletry formulations not intended to have direct contact with mucous membranes.S~“-‘3 Recommended levels are 15 ppm in ‘rinse-off’ and 7.5 ppm in ‘leave-on’ cosmetic products. At present its use for pharmaceutical and dermatological application is not permitted. It has been reported as causing skin irritation and sensitization from industrial exposure, and in patch tests at concentrations > 100 ppm, significantly higher than its level of use in cosmetic products.5v6z’“‘6 Cetyl trimethyl ammonium bromide (CTAB) is a quaternary ammonium compound with a narrow spectrum and microbicidal against Gram-positive bacteria and some Gram-negative bacteria but ineffective against many

Activity

of chlorhexidine,

Kathon

and CTAB

89

Gram-negative bacilli involved in hospital-acquired infections. Quaternary ammonium compounds are cationic detergents which affect the cytoplasmic membrane, causing leakage of cytoplasmic constituents, and also denature and precipitate proteins. l7 Microbial contamination of disinfectant and ammonium compounds has been antiseptic solutions of quaternary implicated as a cause of infections.i8 20

Materials

and

methods

Determination of MIC and MMC The MIC and MMC of chlorhexidine diacetate (Sigma), cetyl trimethyl ammonium bromide (Harcross Chemicals), ‘Kathon’ CG (Rohm & Haas) and two 4% chlorhexidine handwash preparations (‘Bioprep’, Gibson Chemicals Ltd; and ‘Hibiclens’, ICI, Australia) were determined for Staphylococcus aweus NCTC 4163, eight clinical isolates of methicillinresistant S. aweus (MRSA) (collected over three months from one hospital), Escherichia coli NCTC 8196, Proteus vulgaris NCTC 4635, Pseudomonas aeruginosa NCTC 6749, a clinical isolate of Serratia marcescens, a laboratory strain of Candida al&cans and a clinical isolate of Allicrococcus luteus. One-millilitre volumes of two-fold serial dilutions in Tryptone Soya Broth (TSB), Mueller-Hinton Broth (MHB), and Synthetic Broth, AOAC (SB), all Oxoid media, were inoculated with 2-5 x 10” cfu of an 18-h culture of test strains and the MIC was read after 24 and 48 h incubation at 35°C. The MMC was determined by subculture of 2.5 1.11of all concentrations showing no visible growth into 10 ml of TSB containing 3% Tween 80 (BDH Chemicals) and on to Oxoid Columbia Agar Base (CAB) and incubation for 48 h at 35°C. No viable organisms indicated at least a 4 log,,, reduction in inoculum. The accuracy of the determination of surviving cell numbers depends on the adequacy of neutralization of carry-over agent activity and recovery of injured cells. If a sufficiently large inoculum is used (10” or 10’cfu ml-‘) and a 4 log,,, reduction is accepted as an end-point, adequate neutralization is more readily achieved. Both recovery media gavre effective neutralization of carryover activity for all agents for this level of detection. MIC and MMC were also determined in microtitre trays using 200 ~1 volumes and an inoculum five-fold less than for the macro method. Concentrations of ‘Bioprep’ and ‘Hibiclens’ are expressed as concentration of chlorhexidine. Interaction between chlorhexidine and ‘Kathon’ against S. aweus, P. aeruginosa and S. maycescens was evaluated by measuring the 48-h MIC in hIHB, and kill curves in distilled water over an S-h period, of combinations of sub-inhibitory concentrations. ‘I’he effect, on the MIC and MMC in SB, of 10% and 20% horse serum, 0.5% and 2% autoclaved yeast cell suspension, 1% and 5% sterile rabbit faeces and 2% lecithin + 1% Tween 80 on the action of chlorhexidine, cetyl

90

G. Nicoletti

et af.

trimethyl ammonium bromide, ‘Kathon’, ‘ Bioprep’ the full range of test bacteria, was investigated.

and ‘Hibiclens’,

against

Interaction of chlorhexidine with ingredients of media and formulations The bioavailability of chlorhexidine in TSB and MHB was assessed by comparison of the MIC and MMC, in such media and in a synthetic medium (SB), for selected test organisms. At high concentrations (> 256 mg 1-l) chlorhexidine and CTAB formed precipitates in TSB, MHB and SB (in decreasing degree) while ‘Kathon’ showed minimal precipitation in all media. The supernatants and resuspended precipitates from stock solutions of chlorhexidine at 2048, 1024, 512 and 128 mg 1-l in MHB were serially diluted in parallel with the whole suspensions and the MIC and MMC determined for S. aureus and P. aeruginosa. ‘Bioprep’ formulation ingredients identified by serial number (supplied by Gibson Chemicals Ltd) were tested for antimicrobial activity (MIC and MMC). Mixtures of active and inactive ingredients were incorporated into MHB at their relative concentrations in the formulation and tested in parallel with chlorhexidine digluconate (ABM Chemicals) and the complete formulation. Determination of rate of kill The rate of kill of S. aureus, M. luteus, S. marcescens and P. aeruginosa was measured in broth at 35°C and water at 18°C for chlorhexidine, ‘Bioprep’ and ‘Hibiclens’ and in water at 18°C for ‘Kathon’. Bacterial suspensions prepared from 18-h cultures of each strain (2-5 x lo7 cfu mll’ in DW or 2-5 X lo6 cfu ml-’ in MHB) were added to chlorhexidine, ‘Bioprep’, ‘Hibiclens’ or ‘Kathon’. At regular intervals, during 8 h, 100-~1 amounts were transferred into 900-~1 amounts of neutralizing diluent (0*89O/, saline + 3% T80). Survivors were counted by surface spread of 100 ~1, or 3 x 10 ~1 drops of appropriate tenfold dilutions on CAB plates, after 24 h at 35°C. Concentrations were selected to give a 2-5 log,, reduction during 8 h. Adequate neutralization of carry-over activity was not achieved below a 1:lOO dilution for higher concentrations of agent and counts below 100 cfu mll’ were not recorded. The effect of 10% horse serum and 0.5% yeast cell suspensions on the kill rate in water was determined. Time-kill curves were generated by plotting log,, number of survivors against time. Decimal reduction times (DRT) were estimated from the time gradient of the linear regression over a constant time period, over the time to achieve the greatest measurable log reduction or over the linear part of the curve. DRT for agents were plotted against concentration to determine the effect of change in concentration on kill rates. Test reproducibility was evaluated by comparison of the standard deviations of the means, expressed as a percentage of the mean.

Activity

of chlorhexidine,

Kathon

and

CTAB

91

Development of resistance S. aweus, E. coli, S. marcescens, P. aeruginosa and C. albicans were continuously grown for 12 weeks at 35°C in l-ml volumes of MHB containing a range of sub-inhibitory concentrations of chlorhexidine, ‘Hibiclens’ and mixtures of these agents. CTAB, ‘Kathon’, ‘Bioprep’, Subculture of 100 ~1 from the highest concentrations showing light to moderate growth was made every 48 or 72 h to a fresh range of concentrations and MIC and MMC recorded. At intervals the MIC and MMC of chlorhexidine, CTAB and ‘Kathon’ against all strains were measured in both lL’IHB and SB. Resistant strains of P. aeruginosa and S. marcescens were tested for stability by serial subculture, every 48 or 72 h, for 20 passages in drug free MHB during a 7-week period. Cross-resistance of the parent strains and the eight resistant strains from each disinfectant mix, to chlorhexidine, CTAB and ‘Kathon’, was tested in MHB and SB by the micro method for MIC determinations.

Results MIC and MMC The micro and macro methods for the determination of MIC and MMC gave the same results. The difference between MICs and MMCs was greatest at 24 h. MICs at 48 h were usually two to four-fold higher than at 24 h and for all agents and all organisms were more consistently reproducible. MICs and MMCs of the test agents in TSB, MHB and SB at 48 h for a range of organisms are reported in Table I. All MICs and MMCs reported are means (log, scale) of at least four observations and did not differ more than eightfold. MMCs of chlorhexidine, CTAB and ‘Kathon’ were generally 2-8 times higher than MICs in all test media for Gram-negative bacteria and 4-16 times higher for staphylococci, the difference being greater in SB. There was a difference for all strains in the 48 h MICs in the different media. MICs in MHB were often 2-4 times, and in SB 2-16 times lower than those in TSB. ‘Kathon’ showed the least (at most two-fold), and chlorhexidine the most, variation with different media. ‘Kathon’ was the most effective agent against all strains. It was bactericidal for the Gram-negative bacteria and C. albicans and bacteristatic for S. auyeus, the MMC being generally more than 8 times higher than the MIC. Chlorhexidine, ‘Hibiclens’ and ‘Bioprep’ did not differ significantly and vvere generally 4-16 times more active than CTAB. Eight hospital isolates of methicillin-resistant S. aweus were 224 times more resistant than the NCTC strain to chlorhexidine in all media but showed similar Combinations of chlorhexidine and susceptibility to CTAB and ‘Kathon’. ‘Kathon’ at concentrations 4 times lower than their MICs were inhibitory to S. marcescens and P. aeruginosa but not to S. aweus. No synergy in rate of kill in distilled water at 18°C during an 8-h period was demonstrated for

C. albicans

32

64 256

P. vulgaris

i

:

186

64 128

8 16

128

2

4 16

16 64

64 128

S. marcescens

48

128

6

12

32 64

8

:

8

4

iti

16

128 256

128 256

> 1024

64 128

8 32

32 256

2654

P. aeruginosa

0.5 2 16

42

f

0.25 8

TSB

2

:

ii

E. coli

82

SB

(mg 1-l)

16

f

0.5 8

2 8

S. aureus

MH

MIC MMC Chlorhexidine

0.5

TSB

SB

TSB

‘Bioprep’

in Tryptone Soya Broth (TSB), Mueller-Hinton Broth (MH), of ‘Bioprep’ (4% chlorhexidine handwash), chlorhexidine diacetate, ‘Kathon’

MH

Species

(48 h, 35°C)

Table I. Effect ofgrowth

MMC

128 256

128

128

> 1024

512

16

16

16 32

MH

CTAB

128

32 64

2:

128

0.5

0.5

1 1

1

0.25 0.5

1

0.5

1 2

0.5 0.5

4

0.5

MH

1

1

1

2

1

1

i

TSB

‘Kathon’

0.25 0.25

0.12 0.25

0.25 0.5

0.5 1

0.25 0.5

0.12 2

SB

Broth (SB) on the MIC and the ammonium bromide (CTAB) and

!i

16

2

SB

and Synthetic cetyl trimethyl

P z -* 8 ii & z 5

Activity

sub-inhibitory not tested.

of chlorhexidine,

concentrations

Kathon

and

of both agents. Higher

CTAB

concentrations

93

were

EfSect of organic soil Tests with organic soil were more difficult to standardize and less reproducible. Lecithin + Tween 80 had the greatest inactivating effect, generally increasing the MIC by more than 256 times, followed by dried yeast cells (4-128 times), sterile rabbit faeces and horse serum (4-16 times). The inactivating effect of all soil was greatest on microbicidal activity, particularly rate of kill. The higher concentration of yeast cells, faeces and serum were at most twice as inactivating as lower levels and gave less reproducible results. Horse serum (10%) and 0.5% yeast cells gave reproducible results (variation in MIC < 8-fold) and were used as realistic approximations of the level of organic soil likely to be encountered by disinfectants in use. Their effect on the 48 h MICs in SB for a range of bacteria is shown in Table II. Activity was generally reduced more against the staphylococci than against the Gram-negative bacilli. ‘Kathon’ was the agent least affected by all forms of soil, particularly in activity against the Gram-negative bacteria. Yeast had a much greater inactivating effect than serum for the cationic agents but not for ‘Kathon’. CTAB showed less inactivation than the chlorhexidine agents and ‘Bioprep’ than ‘Hibiclens’. Effect of medium and formulation components on chlorhexidine The MICs for S. aureus and P. aeruginosa derived using stock solutions of chlorhexidine in MHB at 2048, 1024, 512 and 128 mg 1-l were the same for the whole suspensions and the supernatants. The resuspended precipitates had no measurable activity. Removal of the precipitate appeared not to have removed chlorhexidine to any extent measurable by a change in MIC. Five ingredients in ‘Bioprep’ had varying antimicrobial activity (measured by MIC and MMC) against some or all of the test strains (data not shown). One was active against all strains and two against the staphylococci and C. albicans (MIC < 64 mg 1-l). Three were moderately active (MIC > 64 mg 1-l) against the latter and E. coli. Eight ingredients had no activity at concentrations up to 10 000 mg 1-l. When chlorhexidine was tested with the group of inactive ingredients there was a twofold decrease in activity against S. aureus and C. albicans. The mixture of active ingredients, excluding chlorhexidine, was as active as chlorhexidine against S. aureus and C. albicans, but 8-64 times less active against the Gram-negative bacteria. The combination of chlorhexidine and the other active ingredients was twice as active as chlorhexidine alone against S. aweus and MRSA and equally active against the Gram-negative bacilli and C. albicans. The complete mixture of active ingredients gave the same MIC as the complete formulation for all test strains, indicating no adverse or potentiating effect on the MIC of chlorhexidine. Both formulations were

0.5

0.5

8

4

S. aureus

E. coli

P. aeruginosa

S. marcescens

Species

32

64

2

4

s

‘Bioprep’

128 32

8 4

64

256

2

0.5

8

16

s 0.5

32

Y

‘Hibiclens’

128

256

32

64

Y

4

8

0.5

32

64

2

4

s

(mg 1-l) Chlorhexidine

0.25

MIC

128

128

32

64

Y

64

128

8

2

Y 64 64 1024 512

s 32 32 1024 512

CTAB

0.25

0.5

0.25

0.12

‘Kathon’

2

4

1

8

s

2

4

1

8

Y

Table II. Effect of 10% horse serum (S) and 0.5% yeast cells (Y) on the MIC of chlorhexidine, ‘Bioprep’ and ’ Hibiclens’ (4% chlorhexidine antiseptic handwashes), cetyl trimethyl ammonium bromide (CTAB) and ‘Kathon’ in Synthetic Broth (48 h 35°C) against S. aureus, E. coli, P. aeruginosa and S. marcescens 9 2: 8-t E b ; F

P

Activity

of chlorhexidine,

Kathon

and

CTAB

stable after 18 months’ storage at room temperature in the dark, showing difference in MIC or kill rate when compared with fresh preparations.

95

no

Rate of kill Time-kill curves were much less reproducible than MIC determinations. Inoculum size and growth stage, medium pH and composition, temperature and cell dispersal before counting were shown to be important factors affecting reproducibility. Results in water were more reproducible than those in broth. Use of log-phase cultures gave faster and more reproducible kill rates (data not shown) but, to simulate better the differing physiological states in natural populations, 16-l 8 h cultures were routinely used. Because these agents were used as controls for testing new agents the concentrations tested (usually a dilution factor) varied and comparisons at equivalent concentrations cannot always be made. Tables III and IV show log reduction factors (RF) of various concentrations of chlorhexidine for S. aweus, P. aeruginosa and S. marcescens in distilled water (18°C) and Mueller-Hinton broth (35°C). Values reported are means of at least three observations. The SD, expressed as a percentage of the mean, varied between 6 and 18% for measurements in water and between 7 and 22% for those in broth. Bacteria in water without disinfectant showed RF < 0.5 over 4 h and in MHB showed increases of > 2 over 8 h. Decimal reduction times (DRT) in water were estimated log,,, from time-kill curves over the time to achieve reduction to < 100 cfu ml-‘. Pearson correlation coefficients varied from 0.76 to 0.97. In broth the DRT were estimated over the linear portion of the curves and Pearson correlation coefficients varied from 0.87 to 0.98. Kill rates were much slower in broth than in water for all strains, a concentration at least 8 times higher being required in MHB to achieve the same level of kill, despite the higher temperature. If rate of kill is compared at similarly effective concentrations (related to respective MIC), chlorhexidine was more rapidly bactericidal to P. aeruginosa and S. maycescens than to 5’. aureus in both water and broth (Tables III and IV, Figures 1 and 2). Chlorhexidine in MHB at concentrations > 8 times the MIC produced a > 5 log,,, reduction for the Gram-negativ,e bacilli in less than one min whereas S. aweus required a concentration of 512 mg ll’ (256 times the MIC) to achieve the same level of kill (data not shown). At concentrations in broth twice their respective MICs P. aeruginosa was more rapidly killed than S. marcescens and both than S. aweus (Figure 1). ,4t concentrations up to 8 times the MIC there was initial rapid kill of susceptible Gram-negative bacilli followed by regrowth of persistent cells within 8 h. Both kill and recovery occurred sooner with P. aeruginosa. Recovery of persistent cells from S. aweus for similar concentration levels above the MIC was not noted within 8 h. The persistent cells, when retested, had the same MIC and MMC as the original inoculum. The same relative activity rate was seen in distilled water (Figure 2) but no attempt

‘Bioprep’

Chlorhexidine

‘Bioprep’

Chlorhexidine

‘Hibiclens’ ‘Bioprep’

Chlorhexidine

‘Hibiclens’ ‘Bioprep’

Chlorhexidine

Agent

Table III. aeruginosa

SM PA SA SM PA SM PA SM PA

8

PA

EI PA SA

Ei PA

SA SM SA

Species

6

4

2

Concentration (mg 1-7

4.4 3.5

1.1

;:g

>5 >5

4.3 3.6

2.5

1.2

0.8

0.7 < 0.5 1.2 1.9

1.1 1.6 1.5 1

5

0.8 1.1 0.5 1.2 1.9 2.6 2.6

0.6
1

3.4 2.2

4.4 >5

2% 1.3 4.1

1.9 1.9 1.1 1.8 2.9 i:i

2.1 1.8

1.2 1 ;:;

Log,, 15

4.1 3.1

>5

3.4 2.7 >5

2.5 2.4 2 3.5 4.6 4.5 4

2.4 2.3

1.6 1.4 ;:;

>5 >5 4.7

;:g

5 4.4

4.3 3.6

>5

2.8 >5

;: 2.5

3:4

2

>5 4,6

4.8 4.2

>5

4.1

4.1

4.9 3.5 3.4

>5 4

180

4.3 3.4

2.9 2.7 >5

reduction factors at min 30 45 60 120

> 5

>5 >5

>5

3.5 3.2

240

%

:: <5 <5

<5 <5

z

;

5 &

z g-

0

12

48 9

28 52

i: 52

E

80

DRT min

S. aureus ISA), S. marcescens (SM) and P. Rate of kill by chlorhexidine, ‘Bioprep’ and ‘Hibiclens’ of 1-3~ lO’cfuml(PA) in distilled water at 18”C, over 4 h, measured by log reduction factors (initial-survivor log,, counts) and estimated decimal reduction times (DRT)

16

32

Chlorhexidine ‘Hibiclens’

Chlorhexidine

‘Hibiclens’

Chlorhexidine

‘Hibiclens’

Chlorhexidine

128

64

6

Chlorhexidine ‘Bioprep’

‘Hibiclens’

4

:itI PA SM

PA SM PA SM

PA SM PA SM

EI

PA

SA

SA

Species

4.4 I.6 >5 >.5

1.8 2 2.7 2.4

CO.5
co.5 co.5
co.5 co.5

CO.5 co.5

om?

2.7

>5

>s >5

2.9 2,2

I.4 I.1 3.8 3.2

co.5 0.5 I.2

< 0.5 co.5

co.5 < 0.5

0.25

>5 4.4

4.1 2.7 >5 >5

2 I.5 4.9 4.1

<0.5 I.1 2.1

co.5 0.6

co.5 CO.5

> 5 >j

>5 3,8 >5 > 5

3.2 2.4 >5 4.8


0.8 1

CO.5 0.5

L og,,, reduction 0.5 1

‘Bioprep’ and ‘Ilibiclens’ (4% chlorhexidine handwashes), (PA) in MHB at 35°C over 8 h measured by log reduction decimal veduction times ( DR T)

Concentration cm!4 1-l)

Chlorhexidine ‘Bioprep’

Agent

Table IY. Rate of kill by chlorhexidine, marcescens (S&I) and P. aeruginosa

>5

>5 >5 >5 >5

3.5 3.1 3.7 > 5

2.3 2.7 4.1



factors, 2

>5

>5 >5 >5 >S

3.6 4.3 3.2 >5

I.8 2.6 3.2

I.5 I.9

I.2 1.6

at hours 4

of I-3 x IO’cfu ml-’ factors [initial-surviror

>5

>5 >5 3.1 >5

3.2 3.4 2.4 3.7

I.3 2 2.9

I.9 2.6

I.3 2.1

6

2-3

3.4 >5 2.7 >s

3 3.5 2.2 3.1


2.3 3.2

1.6 2.5

8

F ii a

18

“g

P ?i. E c%

B 6

4

p 0 c. 5. 4

68

19 26

30 21

211 165

335 183

DRT 8

S. aureus (SA), S. log,,) rounts) and

G. Nicoletti

et al.

200

300

400

500

Time (min) Figure 1. Time-kill curves for chlorhexidine, ‘Bioprep’ and ‘Hibiclens’ at 4 mg 1-l concentration of chlorhexidine for S. aureu~ (SA), and at 32 and 64 mg ll’ for S. marcescens (SM) and P. aeruginosa (PA) in MHB at 35°C. The 48 h MICs in MHB for chlorhexidine are respectively 2, 16 and 32 mg ll’. These are a representative set of counts from those used to obtain average log reduction factors reported in Table IV. PWp Chlor32PA, ~ Chlor32SM, - - - -A- - - Hib32PA, -- 0 -~ Biop4SA, ---•O--Chlor64PA, ~ Chlor64M, - - - - ~5 - - - Hib32SM, PO--Chlor4SA.

was made to detect surviving cells not capable of forming colonies within 24 h on CAB. There was a significant increase in the kill rate of both formulations, compared with chlorhexidine alone, against S. aweus, S. marcescens and P. aeruginosa (Tables III and IV, Figures 1 and 2). The formulations did not differ significantly in rate of activity against the Gram-negative bacilli but ‘Bioprep’ was more rapidly bactericidal than ‘Hibiclens’ to S. aweus and M. Zuteus (Figure 3). Yeast cells and horse serum had a dramatic effect in lowering the rate of bactericidal activity in water of both chlorhexidine formulations (data not shown). Yeast had at least a ten-fold greater inactivating effect than serum. ‘Bioprep’ at 50 mg 1-l produced > 5 log,, reduction of S. aureus in < 1 min in water, a 5 log,, reduction in 6 h in 10% horse serum and < 1 log,, reduction in 6 h in 0.5% yeast cells. Similarly, 64 mg 1-l reduced the population of S. marcescens and P. aeruginosa > 5 log,, within 1 min in water, 2-3 log,, in 8 h in 10% serum and by ~0.5 log,, in 8 h in 0.5% yeast. The reduction in activity against all test organisms achieved by yeast and serum was greater for ‘Hibiclens’.

Activity

of chlorhexidine,

Kathon

so

and

100

99

CTAB

150

Time (min) ITigure 2. Time-kill curves for chlorhexidine and ‘Bioprep’ (both at 4 mg l- ‘) against 5’. aureu~ (SA), P. aeruginosa (PA) and S. mnuescens (SM) in distilled water at 18°C. These are a representative set of counts from those used to determine average log reduction factors reported in Table III. ---m--Bioprep PA, -+Bioprep SA, --A-Chlorhex S%l, ~~~~~~0~~ Bioprep SM, --VPChlorhex PA, ~~C~~ Chlorhex SA.

‘Kathon’ in water was much less bactericidal than chlorhexidine to the test bacteria (Figure 4). It was slowly and v,ariably bactericidal during 8 h for S. UUY~US (RF l-1.5) and S. marcescens (RF 0.8-1.4) for the of concentration range from 1 to 512 mg l-‘, the effect being independent concentration. It was more rapidly bactericidal to P. aeruginosa, the rate of kill increasing with increasing concentration over the range 2-32 mg l- ’ At 32 mg 1-l the population was reduced by 3 log,,, in 8 h. The microbicidal effect over 24. h was much greater against the Gram-negative bacilli with reduction factors of > 5 for P. aeruginosa and S. marcescens compared with 223 for S. aweus. Although S. marcescens was generally less susceptible than S. aUYeUs ov’er 8 h, at 24 h no survivors were detected, probably because of the much poorer survival of Gram-negative bacteria in water. ‘l’he time-kill curves for chlorhexidine agents over 8 h were not linear, departure from linearity being greater for the Gram-negative bacilli than for staphylococci and micrococci (Figures l-3), suggesting more heterogeneous population susceptibility in the former. Time-kill curves for ‘Kathon’ were

100

G. Nicoletti

0

ef al.

100

50

150

Time (min) Figure 3. Time-kill curves for ‘Bioprep’ and ‘Hibiclens’ (1.6mgl-’ expressed as concentration of chlorhexidine) against S. aureu~ (SA) and M. luteus (M) in distilled water at 18°C. Mean counts from three observations. The standard deviation of the mean, expressed as percentage of the mean, varied between 5 and 14%. The MIC and MMC for both agents against M. luteus were respectively 1 and 4 mg 1-l. pm--Bioprep SA, - -+- - Hibicl Bioprep M, --O-Hibicl M. SA, POP

linear over 8 h for all three strains suggesting more homogeneous susceptibility. Decimal reduction times provided a simple, single index for comparison of agent activity rates (Tables III and IV). Their use also permitted a comparison of change in kill rate with concentration (Figure 5). Over the short concentration ranges tested kill rate for chlorhexidine agents is seen to be concentration dependent but not linear, suggesting some limitation on the rate of agent interaction with cells. ‘Kathon’ showed a similar relationship only for P. aeruginosa (Figure 6). Development of resistance Chlorhexidine engendered the greatest level of resistance in all strains and ‘Kathon’ the least. Resistance to chlorhexidine was least with C. albicans and greatest with the Gram-negative bacilli. After 12 weeks’ continuous exposure to chlorhexidine in MHB at 3572, C. albicans showed a four-fold rise and S. aureus and E. coli a gradual rise in MIC of, respectively, 16 and 32 times (data not shown). P. aeruginosa and S. marcescens, when tested against chlorhexidine, ‘Bioprep’, ‘Hibiclens’ and CTAB, developed rapid and significant resistance to the cationic agents within 4 weeks, reaching a maximum at 8 weeks, with exposure to 12 weeks producing no further increase in resistance (Table V). Both strains showed sudden jumps in MICs from 128 or 256 to > 1024 mg l- ‘. S. marcescens developed the

Activity

of chlorhexidine,

100

Kathon

200

300

and CTAB

101

400

Time (min) Figure 4. Time-kill curves for ‘Kathon’ against S. aureu~ (SA) and S. marcescem (SM) at 512mgl-’ and against P. aevuginosa (PA) at 2, 4, 8, 16 and 32 mgl-‘, in distilled water at 18°C. iLlean log counts from at least four observations. The standard deviation of the mean, expressed as a percentage of the mean, varied between 7 and 20%. -p-m-pPA2, PAP PA32, p-O-SA 512, --Up PA4, -0-m PA16,

greatest level of resistance to all agents. Both strains developed the greatest resistance to chlorhexidine, the MIC increasing 32 times for P. aeruginosa and 128 times for S. marcescens, compared with 16 and 32 times, respectively, for CTAB and twice and 8 times for ‘Kathon’. Resistance levels developed to ‘Bioprep’ were similar to those to chlorhexidine. P. aeruginosa developed a four-fold lower level of resistance to ‘Hibiclens’. An equipotent mixture of chlorhexidine and CTAB was synergistic, both in activity level and in suppression of the development of resistance (Table V). The initial MIC of the mixture was, respectively, 4 and 128 times lower than that for either agent alone against both strains and the final MIC respectively 16 and 128 times lower for P. aeruginosa and 16 and 32 times lower for S. marcescens. The addition of a very low level of ‘Kathon’ (0.04 mg 1-l) further reduced by two-fold the initial and final MIC of both strains. The mixture of chlorhexidine and 0.04 mg 1-l of ‘Kathon’ was only half as active as the mixture of chlorhexidine and CTAB and equally effective in suppressing resistance. The MICs, for all eight resistant strains, of chlorhexidine, CTAB and ‘Kathon’ in MHB and SB, before and after 20 passages in drug-free

102

G. Nicoletti

et al.

J 8

6

4

2

Chlorhexidine

10

pg ml-’

Figure 5. Effect of concentration on the decimal reduction time (DRT) of chlorhexidine and ‘Bioprep’ for S. aureus (SA), P. aeruginosa (PA) and S. marcescens (SM) in distilled water at 18°C. DRT estimated over the time to reduce the inoculum to < 100 cfu ml-‘. Mean of at least four measurements. The standard deviation of the mean, expressed as a percentage of Chlorhex PA, the mean, varied between 9 and 22%. --¤-Bioprep SM, -+Chlorhex SM, --O-Bioprep PA. --o-

310 290 270 250

190 170

I 5

I

I

10

15

20

25

1 30

35

‘Kathon’ pg ml-’ Figure 6. Effect of concentration on the decimal reduction time (DRT) of ‘Kathon’ for P. aeruginosa in distilled water at 18°C. DRT estimated over time to reduce the inoculum to < 100 cfu ml-‘. Mean of four measurements. The standard deviation of the mean DRT, expressed as a percentage of the mean, varied between 12 and 21%.

Activity

of chlorhexidine,

Kathon

103

and CTAB

Table V. Effect on the MIC of disinfectants and disinfectant mixtures, determined in Muellerafter 8 Hinton broth (MHB) and Synthetic Broth (SB), for P. aeruginosa and S. marcescens weeks continuous exposure to disinfectants in Mueller-Hinton broth at 35°C MIC (mg I-‘), P. aeruginosA .’

.4gent Initial

Final

MHB,

SB. 48 h. 35°C S. marcescens

x-fold increase

Initial

s-fold

Final

increase

‘Bioprep’

32 8

1024 128

32 16

16 -I

2048 256

128 64

‘Ilibiclens’

32 8

256 6-1

8 8

16 J

2048 128

128 32

Chlorhexidine

32 x

1024 6-1

32 8

16 I

2048 12X

128 32

512 128

8192 256

16 2

128 6-l

4096 256

32 4

2 I

4 2

2 2

4 I

8 4

8

64

8

4

128

32

Chlorhexidine + ‘Kathon’ (0.04 mg I ‘)

16

128

8

8

256

32

Chlorhexidine + CTAB + ‘Kathon’ (0.04 mg 1-l)

4

32

8

2

64

32

CTAH ‘Kathon’ Chlorhexidine + CTAB (equal concentration)

0.5 (I.25

medium, are shown in Table VI. Lower increases in resistance to all agents were seen if MICs were measured in SB, the difference being greater where MICs were high. The resistant strains developed in chlorhexidine and the antiseptic formulations showed increased resistance to CTAB, twice to 8 times for P. aeruginosa and 8 to 16 times for S. marcescens, when MICs were measured in MHB, but up to 4 times less when measured in SB. Resistant strains developed in the mixtures showed lower increases in resistance to individual agents. The strains resistant to CTAB were normally susceptible to chlorhexidine and ‘Kathon’. For all resistant strains, except those developed in ‘Kathon’, there was a twice to four-fold increase in susceptibility to ‘Kathon’. The ‘Kathon’-resistant strains showed no increased resistance to either chlorhexidine or CTAB. After serial subculture in drug-free medium strains of P. aeruginosa generally retained resistance to chlorhexidine and CTAB but lost the low-level resistance to ‘Kathon’. Strains of S. marcescens were frequently twice to 4 times less

128 256 64

64 128 32

+ ‘Kathon’

4

4

+ CTAB

Chlorhexidine

strain

+ ‘Kathon’

Chlorhexidine

Parent

+ CTAB

Chlorhexidine

‘Kathon’

4096

8192

CTAB

2048

1024

2048

256

‘Hibiclens’

Chlorhexidine

2048

32

32

8

64

64

128

32

64 16

16

128

32

128

32 16

32

16 4

64 16

16

128

64

512

32 8

8

64

8

32

2048

128

512

128

512

128

64

1024

256 64

256 1024 128

512 64 256

256

20

MIC

16

64

64

512

32

32

1024

SM

20

1

1

PA

PA

SM

Chlorhexidine

8

Disinfectant MIC after weeks

1024

to

‘Bioprep’

Resistant

strain-exposure

PA

20 1

SM

128

512

128

512

128

1024

256 64

128

512

512

512

512

256

512

64

128

64

128

64

2.56

256

512

64

128

4096 4096 4096 256 256

1024 1024 1024 128 256

128

128

256

512

128

256

512

2048 1024 2048 1024 256 512

512

20

PA

SM

0.5

1

0.25

0.5

0.25

0.5

42

2

0.5

1

0.5

1

0.25

0.5

0.25 0.25

A.25

0.12

0.5

i

0.12

0.5

0.25

0.5

0.5

0.12

0.5

0.12

0.5

1

0.12 0.5

0.5

0.5

0.5

20

‘Kathon’

0.25

0.5

0.25

0.5

0.25

0.5

1

and SB, 48 h, 35°C

CTAB

MHB

4096 4096 2048 256 512

1

(mg l-l),

0.5

0.25

1

0.5

0.5

0.5

0.5

0.5

0.5

20

VI. MICs of chlorhexidine, cetyl trimethyl ammonium bromide (CTAB) and ‘Kathon’ in Mueller-Hinton (MHB) and Synthetic Broth (SB) for the resistant strains of P. aeruginosa (PA) and S. marcescens (SM) developed after 8 weeks’ continuous exposure to the agents reported in Table III, before and after 20 subcultures in drug free MHB over 7 weeks

Table

%

I;, 2 i;’ g 9 =* %

ii

Activity

resistant after passage chlorhexidine and CTAB.

of chlorhexidine,

but

still

Kathon

retained

and

105

CTAB

significant

resistance

to

Discussion

Chlorhexidine was more broadly active than CTAB but showed more variability in MIC with different media, more inactivation by organic soil and produced a greater development of resistance in Gram-negative bacilli, which also showed increased resistance to CTAB. Although chlorhexidine was rapidly microbicidal at low concentrations the rate of kill was greatly reduced by the presence of organic soil, indicating that it is more suited to use as an antiseptic, or to disinfection where soil levels are low. It can be successfully incorporated into formulations without loss of activity. The chlorhexidine handwashes showed the MIC expected for the theoretical concentration of chlorhexidine present. However, when rate of kill was used as a measure of effectiveness, the antiseptic handwashes were considerably more effective than the equivalent concentration of chlorhexidine alone, showing the importance of formulation components to agent activity. This was also illustrated by small differences in the activities of the two handwashes in rate of kill, activity in the presence of soil and suppression of the development of resistance. Without knowledge of the composition of the two commercial handwash formulations it is not possible to interpret differences in their performance. In the case of ‘Bioprep’, those ingredients having antibacterial activity produced no significant enhancement in the level of activity over that of chlorhexidine but could be responsible for the greater speed of kill. ‘Kathon’ showed a high level of activity against Gram-negative bacilli, a low level of inactivation by media components and organic soil, a low level of induced resistance in P. aeruginosa and S. marcescens and increased activity against chlorhexidine and CTAB resistant strains of these organisms. The lowest effective concentrations were as effective as higher concentrations, confirming its suitability as a preservative. The slow bactericidal rate demonstrated against S. aureus and S. marcescens in water suggests that it would be of limited use as a disinfectant where a rapidly lethal action was required. The difference in killing kinetics for P. aeruginosa and S. marcescens, despite the similarity in MIC, suggests a difference in the mechanism of action. Collier’ reported that the component isothiazolinones in ‘Kathon’ were not bactericidal to E. coli over 24 h at concentrations up to five times the MICs. At very low levels it was significantly synergistic with chlorhexidine. This suggests its use at levels up to 7.5mgl -’ in formulations might be beneficial in potentiating the effect of chlorhexidine and reducing the selection of resistant strains. The importance of standardization of media and sampling time in evaluation of MIC and MMC has been confirmed. At high concentrations the level of free agent in complex media or formulations may be significantly less than the theoretical level, resulting in the reporting of

106

G. Nicoletti

et al.

MICs greater than actual. The use of chemically defined media in all tests is therefore recommended, particularly where MICs are high, to minimize variability due to interaction between agent and medium components, to investigate chemical modulation of activity and to monitor development of resistance. The greater activity of all agents in the synthetic medium used could be due to poorer growth and/or to less sequestration or inactivation of agent by medium constituents. Assessment of the bio-availability in a medium is necessary if activity at high concentrations and increases in bacterial resistance are to be accurately monitored. Such information would also be useful in determining the level of agent to be incorporated into formulations, the effect of dilution on activity and the evaluation of the level of free agent in the case of contamination with ‘resistant’ organisms. Variability in MIC and MMC in different media correlated with response to organic soil. ‘Kathon’ showed the least variability in MIC and the least inactivation by all forms of soil and chlorhexidine the most. Organic soil had a greater effect on the kill rate and the MMC than on the MIC. The effect on kill rate might better predict its effect on agent activity in use. Simulation of environmental conditions is important in assessment of disinfectant efficiency. Selection of standardized soil conditions will depend on the conditions of use. Investigation of the kinetics of disinfection was helpful in differentiating microbistatic from microbicidal activity, differentiating agent effects on different bacteria and showing the effects of dilution and inhibitory conditions on activity. Expression of kill rates by reduction factors over time is simple and informative. Generation of a kill curve characterizes the population response. Estimation of decimal reduction times provides a single index of kill rate and is valuable in assessing the time required to achieve a given level of disinfection by an agent. DRT estimations are based on an assumption of linearity in the relationship between survivor log counts and time. With parabolic relationships, owing to variability in cell susceptibility, the DRT calculated will depend on the portion of the curve selected for estimation. The time to achieve a 4 or 5 log,, reduction may be an appropriate interval. This could overestimate the DRT and provide a safety factor. Estimations from the initial linear part of the curves, although useful in comparing agent activity in vitro, would overestimate the ability to kill in situ. Correlation of in-vitro test results with in-use performance is important for validating the former as predictive tests of efficacy. The faster kill rate by in ‘Bioprep’ and ‘Hibiclens’ of S. marcescens than M. luteus demonstrated vitro was confirmed in a handwash trial using artificially contaminated skin on the hands of volunteers. *’ ‘Bioprep’ and ‘Hibiclens’ produced a greater reduction in the counts of S. marcescens than M. Zuteus in 30 min from the fingertips. The faster in-vitro kill rate of ‘Bioprep’ in water and broth for staphylococci and micrococci was not reflected in vivo. ‘Hibiclens’ showed a greater initial bactericidal action at 2 min, with no increase in effect after 30 min, whereas ‘Bioprep’ exerted a lesser initial effect but a greater

Activity

of chlorhexidine,

Kathon

and

CTAB

107

sustained effect. This may be a reflection of the greater ability of ‘Bioprep’ to withstand inactivation by organic soil. The study showed that individuals varied in their response to the contaminating organism, there being a significant difference between the counts of the two organisms on different people, and to the agent formulation, different agents being more effective in reducing counts for different people. Such effects were not predictable from the in-vitro testing. There have been many reports of the contamination of disinfectant and antiseptic solutions with strains of Gram-negative bacilli, particularly of the genera Proteus, Providencia, Pseudomonas and Serratia, which have become resistant ammonium to chlorhexidine, quaternary compounds or hexachlorophene,‘x 20,22~23 and of the occurrence of disinfectant resistance in clinical isolates.“~‘” L ome S level reports of high ‘resistance’ in Gram-negative bacilli may be spurious. In-vitro determinations of the MIC can overestimate the level of resistance because of lowered bio-availability of the agent. Walker and Lowes *’ found the MIC test to be unreliable in the assessment of chlorhexidine resistance as it was dependent on medium used, inoculum size and age of culture. Meakin 27 found that the ingredients and method of formulation could affect the physical and biological properties of antimicrobial agents. Th e results of this study suggest that lowered availability of free agent at higher concentrations is the cause of high measured ~IICS in MHB. The lovver increases to in resistance chlorhexidine and C’I’AB by P. aeruginosa and S. maYcescerrs when measured by MIC in SB and the smaller difference at lovver concentrations betvveen MIC values in SB and MHB suggest that the concentration of available chlorhexidine or CTAB in the latter at higher concentrations \vas probably progressively less than the theoretical concentration. More valid comparisons of development of resistance can be made by considering relative increases in hIIC in the simplest synthetic medium supporting growth of the strain. 1,owered bioavailability may explain some reported isolations of resistant Gram-negati1.e bacilli from antiseptic solutions. hIarrie and Costerton,‘3 using an agar dilution method, reported isolates of S. maycescens havring ~IICS of 1024 mg ll’, but able to survive in a theoretical concentration of 20 000 mg 1-l in a chlorhexidine formulation (‘Hibitane’). They showed that the organisms were embedded in a fibrous matrix and resistance was probably due to the inaccessibility of the colonizing organisms to the agent. Use of concentrations above an optimum level may not achieve greater control or prevent the emergence of resistant populations, e.g. the use of 4% chlorhexidine in disinfectant formulations may be no more effective than 1% or 2%. The activity level and the development of resistance vvas considerably less when mixtures of disinfectants were used. Testing of possible combinations of agents for improved in-use performance would be important for the control of Gram-negative bacilli. The use of disinfectant mixtures might result in greater activity and less likelihood of emergence of resistant strains and allow the use of lower concentrations of single agents.

108

G. Nicoletti

et al.

Although the transfer of multiple antibiotic resistance by plasmids is widespread among clinically important organisms there is no strong evidence that disinfectant resistance in Gram-negative bacilli is plasmid borne or transferred, or associated with antibiotic resistance.2s31 Plasmid-mediated resistance in Gram-positive and Gram-negative bacteria ions, to divalent metal organo-mercurials, silver-sulphadiazine combinations and hexachlorophene has been demonstrated but does not appear to be associated with resistance to other disinfectants.29’32 Various levels of permanent and transitory resistance to high levels of chlorhexidine have been reported with Gram-negative genera such as Pseudomonas, Serratia and Proteus.26,30,33,34Th e strains reported by Walker and Lowes26 lost resistance on subculture, whereas we observed development of stable resistance by P. aeruginosa and S. marcescens to chlorhexidine and CTAB. Kill curves in MHB demonstrated recovery of persistent cells within 8 h at concentrations of chlorhexidine up to eight-fold greater than the MICs. Stickler3’ showed correlation of resistance to chlorhexidine with resistance to cetrimide and benzalkonium chloride in Proteus mirabilis. The chlorhexidine resistant strains of P. aeruginosa and S. marcescens reported here were cross-resistant to CTAB whereas the strains resistant to CTAB were normally susceptible to chlorhexidine. Plasmid-mediated low-level resistance to chlorhexidine has been demonstrated in staphylococci3’j and strains with increased resistance to quaternary ammonium compounds, chlorhexidine, hexachlorophene, propamidine, povidone iodine and have been reported.28a37*38 acriflavine Brumfitt, Dixon and Hamilton-Miller38 reported MICs of chlorhexidine four times higher for methicillin-resistant S. aureus strains. Our methicillin-resistant S. aureus isolates were two to four-fold more resistant to chlorhexidine and CTAB but equally susceptible to ‘Kathon’. Most recommended normal use dilutions of disinfectants are adequate to suppress phenotypically resistant strains, unless the latter are protected from contact with the agent or the agent is sequestered by chemical components in the environment. However, the demonstration of stable resistance to high concentrations of chlorhexidine and cetyl trimethyl ammonium bromide in this study would suggest that the use of cationic disinfectants and antiseptics in a hospital environment could select for resistant populations. Stickler and Thomas25z39 found many multiply antibiotic resistant strains of P. aeruginosa, P. mirabilis and Providencia stuartii were resistant to chlorhexidine and cetrimide and suggested that reliance on cationic antiseptics was likely to select for antibiotic resistant species. These findings suggest it would be appropriate to monitor genera such as Proteus, Providencia, Pseudomonas and Serratia in the hospital environment for changes in susceptibility to disinfectants. There are few internationally accepted standard methods for the assessment of disinfectant activity 4o in vitro or in situ and the choice of test method is largely dictated by the circumstances of use and the activities Current in-vitro methods use single laboratory or being evaluated.

Activity

of chlorhexidine,

Kathon

and

109

CTAB

under optimum conditions and environmental strains in pure culture, subjected to an agent in a sterile, controlled system. They measure survival of cells capable of growing on routine media within a short incubation time. Such tests as those reported here can give accurate and reproducible results and are valuable for comparative in-vitro evaluation of agents. Activity under these artificial conditions, however, may have little similarity to the complex natural environment where microorganisms are present in mixed populations, in varying physiological states, aggregated and protected, and in varying environments containing different interfering substances. All these factors will greatly affect disinfectant performance. There is a need to develop methods concerned with estimating survival of microorganisms exposed to disinfectants under field conditions. A greater understanding is needed of the interactions of the physical, chemical and microbiological components in the environment of use. Factors needing investigation include strain variation in susceptibility, the behaviour of organisms in mixed populations and different situations and, particularly, the detection and recovery of stressed, starving and injured cells which are unable to form colonies on routine laboratory culture. The latter may be refractory to disinfectant action because of low metabolic activity but still able to infect. In many environmental situations they may be the predominant organisms present but they are ignored in routine testing procedures. The authors ‘Bioprep’.

wish

to thank

Gibson

Chemicals

Ltd

for

the supply

of the ingredients

of

References mechanisms. In: Russell AD, Hugo, WB, Ayliffe GAJ, Eds. 1. Hugo WB. Disinfection Principles and Practice of Disinfection, Preseraation and Sterilization. Oxford: Blackwell Scientific Publications 1982; 158-l 85, antibacterial action and bacterial resistance. Infection 1986; 2. Russell AD. Chlorhexidine: 14: 212-215. U.S. Patent 3 761 488. 25 3. Lewis SN. Miller GA, Low AB. To Rohm & Haas Comnanv. September 1973. Kathon 886 h’Iicrobicide. Broad spectrum antimicrobial 4. Rohm & Haas Company. activity. Technical Bulletin 1983: CS 477. Haas Company. Kathin CG microbicide. Cosmetics and toiletries. Technical 5. Rohm-& Bulletin 1987: CS 536a. JW. Kathon CC: a review. j’ Am Acad Dermatol 1988; 18: 6. De Groot AC, Weyland 350-358. inhibitory and biocidal activity of 7. Collier PJ, Ramsey AJ, Austin P, Gilbert P. Growth some isothiazolone biocides. J Appl Bacterial 1990; 69: 569-577. SF, Brackett CR, Fisher JD. Fates of microbicidal 3-isothiazolone 8. Krzeminski compounds in the environment: modes and rates of dissipati0n.J Agric Food Chem 1975; 23: 1060-1068. SF, Brackett CR, Fisher JD, Spinnler JF. Fate of microbicidal 9. Krzeminski 3-isothiazolone compounds in the environment: products of degradation. J Agric Food Chem 1975; 23: 1068-1075. 10. Monte WC, Ashoor SH, Lewis BJ. Mutagenicity of two non-formaldehyde-forming antimicrobial agents. Food Chem Toxicol 1983; 21: 6955697.

110

G. Nicoletti

et al.

11. Decker RL Jr. Frequency of preservative use as disclosed to FDA-1984. Cosm Toilet 1985; 100: 65-68. 12. Benassi CA, Rettore A, Semenzato A, Better0 A, Cerini R. Chemical stability and microbiological reply of preservative systems in cosmetics: a direct correlation. Int r Cosm Sci 1988; 10: 231-239. 13. Rastogi SC. Kathon CG and cosmetic products. Contact Dermatitis 1990; 22: 155-160. 14. Maibach HI. Diagnostic patch test concentration for Kathon CG. Contact Dermatitis 1985; 13: 242-245. 15. De Groot AC, Herxheimer A. Isothiazolinone preservative. Cause of a continuing epidemic of cosmetic dermatitis. Lancet 1989; 1: 314-316. 16. Reitschel RL, Nethercott JR, Emmett EA et al. Methylchloroisothiazolinonemethylisothiazolinone reactions in patients screened for vehicle and preservative hypersensitivity. J Am Acad Dermatol 1990; 22: 734-738. 17. Petrocci AN. Surface active agents: quaternary ammonium compounds. In: SS Block, Ed. Sterilization, Disinfection and Preservation. Philadelphia, PA: Lea & Febiger 1983; 309-329. 18. Adair FW, Geftic SG, Gelzer J. Resistance of Pseudomonas to quaternary ammonium compounds. I. Growth in benzalkonium chloride solutions. Appl Microbial 1969; 18: 299-302. 19. Dixon RE, Kaslow RA, Mackel DC, Fulkerson CC, Mallison GF. Aqueous quaternary ammonium antiseptics and disinfectants. Use and misuse. J Am Med Assoc 1976; 236: 2415-2417. MJ, 20. Frank Schaffner W. Contaminated aqueous benzalkonium chloride-an unnecessary hospital infection hazard. J Am Med Assoc 1976; 236: 2418-2419. 21. Nicoletti G, Boghossian V, Borland R. Hygienic hand-disinfection: a comparative study with chlorhexidine detergents and soap. J Hasp Infect 1990; 15: 323-337. A. Survival of Gram negative bacilli and Cundidu ulbicuns in 22. Brun JN, Digranes hexachlorophane preparations and other disinfectants. ScandJInfDis 1971; 3: 235-238. 23. Marrie TJ, Costerton JW. Prolonged survival of Serrutiu murcescens in chlorhexidine. Appl Environ Microbial 1981; 42: 1093-I 102. 24. Nakahara H, Kozukue H. Isolation of chlorhexidine resistant Pseudomonas aeruginosa from clinical lesions. J Clin Microbial 1982; 15: 166-168. 25. Stickler DJ, Thomas B. Antiseptic and antibiotic resistance in Gram-negative bacteria causing urinary tract infection. J Clin Puthol 1980; 33: 288-296. 26. Walker EM, Lowes JA. An investigation into in-vitro methods for the detection of chlorhexidine resistance. J Hasp Infect 1985; 6: 389-397. BJ. Some physico-chemical factors affecting the antimicrobial activity of 27. Meakin antiseptic formulations. Br J Clin Pratt (Suppl. 125), 1983; 12-22. 28. Ahonkai I, Russell AD. Response of RPI+ and RPIstrains of Escherichia coli to antibacterial agents and transfer of resistance to Pseudomonas uernginosu. Cnrr Microbial 1979; 3: 89-94. 29. Russell AD. The role of plasmids in bacterial resistance to antiseptic disinfectants and preservatives. J Hosp Infect 1985; 6: 9-19. 30. Michael-Briand Y, Laporte JM, Bassignot A, Plesiat P. A note on antibiotic resistance plasmids and bactericidal effects of chlorhexidine on Enterobacteriaceae. J Appl Bact 1986; 61: 106-111. 31. Russell AD, Hammond SA, Morgan JR. Bacterial resistance to antiseptics and disinfectants. J Hosp Infect 1986; 7: 213-22.5. I, Pugh WJ, Russell AD. Sensitivity to antimicrobial agents of some mercury 32. Ahonkhai sensitive and mercury resistant strains of gram negative bacteria. Curr Microbial 1984; 11: 183-185. 33. Prince NH, Nonemaker WS, Norgard RC, Prince DL. Drug resistance with topical antiseptics. J Pharmaceut Sci 1978; 67: 1629-I 63 1. 34. Stickler DJ, Thomas B, Clayton CL, Chawla JC. Studies on the genetic basis of chlorhexidine resistance. Br J Clin Pratt 1983; Symposium no. 25: 23330. 35. Stickler DJ. Chlorhexidine resistance in Proteus mirabilis. J Clin Puthol 1974; 27: 284-287. 36. Townsend DE, Ashdown N, Greed LC, Grubb NB. Transposition of gentamicin resistance to staphylococcal plasmids encoding resistance to cationic agents. Antimicrob Chemother 1984; 14: 115-124. 37. Townsend DE, Greed L, Ashdown N, Grubb WB. Plasmid-mediated resistance to

Activity

of chlorhexidine,

Kathon

and CTAB

111

quaternary ammonium compounds in methicillin-resistant Staphylococcus aureus. MedJ Aus 1983; 2: 310. 38. Brumfitt W, Dixon S, Hamilton-Miller JMT. Resistance to antiseptics in methicillin and gentamicin resistant Staphylococcus aureus. Lancet 1985; 1: 1442-1443. 39. Stickler DJ, Thomas B. Intrinsic resistance to non-antibiotic antibacterial agents. In: Russell AD, Hugo WB, Ayliffe GAJ, Eds. Principles and Practice of Disinfection, Pvesereation and Sterilization, Oxford: Blackwell Scientific Publications 1982; 1866198. 40. Reybrouck G. International standardization of disinfectant testing: is it possible?JIJosp Infect 1991; 18 (Suppl. A): 28(t288.