Journal of Antimicrobial Chemotherapy (2000) 46, 551-556
© 2000 The British Society for Antimicrobial Chemotherapy
In vitro activity of new generation fluoroquinolones against genotypically distinct and indistinguishable Clostridium difficile isolates
Department of Microbiology, University of Leeds and The General Infirmary, Leeds LS2 9JT, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
We compared the activities of ciprofloxacin and levofloxacin with those of the newer fluoroquinolones grepafloxacin, moxifloxacin, sparfloxacin and trovafloxacin against Clostridium difficile isolates. As there is good evidence of marked clonal spread of C. difficile, we studied both genotypically distinct (n = 26) and indistinguishable (n = 28) isolates as determined by random amplified polymorphic DNA and ribosomal spacer PCR fingerprinting. The indistinguishable strains examined represent the main UK epidemic C. difficile clone. For 17 of 54 strains (31%) we were unable to read MICs following inocula preparation using Mueller–Hinton broth. Using Schaedler's broth for inocula preparation 93% of strains had readable MICs, although geometric mean MICs were uniformly higher (2.5- to 5.4-fold) compared with results using Mueller–Hinton broth. Moxifloxacin and trovafloxacin, followed by grepafloxacin, were the most active fluoroquinolones tested and were 3- to 4-fold more active than older agents such as ciprofloxacin by both MIC methods. Unexpectedly, clonal C. difficile strains had markedly reduced susceptibility compared with the distinct strains to each of the fluoroquinolones tested. Clonal strains were more than seven-fold or 12- to 29-fold less susceptible (according to geometric mean MICs) than distinct strains to both moxifloxacin and trovafloxacin, depending on the MIC method used. It remains to be seen whether the enhanced activity of new fluoroquinolones such as moxifloxacin in comparison with other fluoroquinolones against C. difficile implies that these agents are unlikely to be associated with C. difficile infection. However, clinical use of new generation fluoroquinolones in elderly hospitalized patients where C. difficile is endemic requires further study, particularly given the reduced antibiotic susceptibility to all fluoroquinolones of the readily transmissible UK C. difficile clone.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Reports of Clostridium difficile continue to increase markedly in the UK, with evidence for major disruption of hospital activity and marked cost implications.1–3 Numerous studies have cited second- and third-generation cephalosporins as significant risk factors for C. difficile infection.4–6 By contrast, aminoglycosides, first-generation fluoroquinolones such as ciprofloxacin and ureidopenicillins with or without ß-lactamase inhibitors are associated rarely with C. difficile diarrhoea.5–7 Poor anti-anaerobic activity by aminoglycosides and first-generation fluoroquinolones, and hence the preservation of resistance by gut flora to colonization by C. difficile, is believed to explain at least in part the relative protection afforded by these antimicrobial agents.
New generation fluoroquinolones have improved activity against both Gram-positive bacteria and anaerobes.8 It is therefore uncertain whether these antimicrobial agents will share the low propensity of progenitor antibiotics such as ciprofloxacin to cause C. difficile infection. There are few published data available to determine the comparative activity of new generation fluoroquinolones against C. difficile.9,10 Furthermore, there is evidence that there is much clonal spread of C. difficile strains, and for example, 55% of C. difficile strains submitted to the Public Health Laboratory Service Anaerobe Reference Unit by UK hospitals are indistinguishable genotypically.11 In Leeds General Infirmary this epidemic strain causes c.80% of symptomatic C. difficile infections in elderly patients (unpublished data). Despite these observations, in vitro antibiotic activity studies are performed commonly on C. difficile isolates that have not been fingerprinted to ensure that varied strains are selected for study. This situation is analogous to the selection of MRSA isolates for study without regard for the fact that very limited numbers of DNA clones are responsible for the great majority of infections.12
We therefore aimed to compare the activities of ciprofloxacin and levofloxacin with those of the newer fluoroquinolones grepafloxacin, moxifloxacin, sparfloxacin and trovafloxacin against both genotypically distinct and indistinguishable C. difficile isolates.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
We screened a total of 220 C. difficile clinical and hospital environmental isolates by random amplified polymorphic DNA (RAPD) and ribosomal spacer PCR, and selected 26 genotypically distinct strains (more than three bands difference in DNA profile) plus 28 clonal strains, which were obtained from specimens separated in time and space (Figure 1
). The DNA clone is a UK epidemic strain designated PCR ribotype 1 by Brazier and colleagues.11 The MICs of ciprofloxacin, grepafloxacin, levofloxacin, moxifloxacin, sparfloxacin and trovafloxacin (each obtained from Bayer, Newbury, UK) were determined by two agar incorporation methods. For method 1, according to NCCLS guidelines,13 bacteria were cultured anaerobically in Mueller–Hinton broth (Merck, Darmstadt, Germany) at 37°C for 48 h, and then strains were multipoint-inoculated (104 cfu/spot, adjusted using a MacFarland standard tube) on to Wilkins–Chalgren agar (Oxoid, Basingstoke, UK) containing doubling dilutions of antibiotic (0.03–128 mg/L). For method 2, we used Schaedler's anaerobic broth (Oxoid) instead of Mueller–Hinton broth for inocula preparation with all other conditions the same as for the first method. All media and diluents were pre-reduced in an anaerobic cabinet for at least 2 h. For both methods MICs were read after culture at 37°C anaerobically for 48 h. Bacteroides fragilis NCTC 6343, Staphylococcus aureus NCTC 6571 and Escherichia coli NCTC 10418 were used as control strains. The MIC was defined as the lowest concentration that prevented visible growth.
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The MIC results for the genotypically distinct and the clonal strains are shown in Tables I and II
(methods 1 and 2, respectively). With method 2, as opposed to method 1, the great majority of strains (93% versus 69%) yielded readable MICs. A total of 30 strains (20 clonal and 10 distinct) yielded readable MICs by both methods, and these were markedly higher (geometric means 2.5- to 5.4-fold greater) when method 2 was used.
|
|
Moxifloxacin and trovafloxacin, followed by grepafloxacin, were the most active fluoroquinolones tested. In general, these newer fluoroquinolones were 3- to 4-fold more active than older agents such as ciprofloxacin, by both MIC methods. As expected there was less variability in the MICs of all antibiotics tested for the clonal as compared with the genotypically distinct strains. However, an unexpected finding was the markedly reduced susceptibility of the clonal C. difficile strains to each of the fluoroquinolones compared with the distinct strains. For example, clonal strains were more than seven-fold less susceptible than distinct strains to both moxifloxacin and trovafloxacin (according to geometric mean MICs) using method 1. This difference in susceptibility was even greater when method 2 was used (12- to 29-fold). This phenomenon is illustrated in Figure 2a and b
, where the moxifloxacin and ciprofloxacin MIC results for the clonal strains are clustered to the right-hand end of the x-axis.
|
) Clostridium difficile strains, according to method 2. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
This is the first study of antibiotic activity against C. difficile to examine both clonal and distinct strains. Unless a discriminatory fingerprinting method is employed, collections of clinical isolates may be severely biased by the local epidemiology of C. difficile. Stubbs et al.11 reported recently their experience with DNA fingerprinting C. difficile isolates submitted to an Anaerobe Reference Unit. They identified 116 distinct types on the basis of differences in DNA profiles generated with PCR primers for the 16S–23S rRNA gene intergenic spacer region. However, isolates from 55% of infections in UK hospitals belonged to one ribotype (type 1), although this type was responsible for only 7.5% of community infections. PCR ribotype 1 causes c.80% of C. difficile infections in this hospital and therefore the great majority of isolates selected at random from clinical specimens can be expected to be clonal. Similarly, Rafferty et al.14 found that 53–84% of C. difficile clinical isolates, collected over two periods 2 years apart, were indistinguishable depending on the fingerprinting method used. Conversely, a study of an oncology ward showed that while 29% of environmental strains were indistinguishable, 18/21 C. difficile patient isolates represented distinct groups.15 Hence, the constitution of a collection of C. difficile isolates, in terms of the proportion of clonal strains, cannot be gauged without the use of appropriate discriminatory fingerprinting.
Other workers have not addressed the issue of clonal versus distinct strains when selecting study isolates, and this may reflect the fact that potentially epidemic C. difficile strains have only recently been recognized. Interestingly, in three studies that have examined the activity of fluoroquinolones against C. difficile, MIC90s were found to be lower than those observed in the present study for clonal strains.9,10,16 Fluoroquinolone MICs for genotypically distinct C. difficile strains observed in the present study using conventional inocula preparation are similar to those reported by Nord (ciprofloxacin)16 and Woodcock et al.10 (ciprofloxacin and moxifloxacin). Edlund et al.9 reported an MIC90 of moxifloxacin for 50 C. difficile isolates (of uncertain relationship) of 2 mg/L. While the corresponding figure in the present study was 16 mg/L, the geometric mean MIC90 was only 1.3 mg/L, indicating that the MIC was elevated by a small number of relatively resistant strains.
We opted to use two different methods to prepare inocula for MIC determinations in an attempt to overcome the often poor growth that was obtained. For 17 of 54 strains (31%) we were unable to read MICs (i.e. no growth on the antibiotic-free control plate) following inocula preparation using Mueller–Hinton broth. Hence, we used instead Schaedler's anaerobic broth to prepare inocula for MIC determinations. Schaedler's broth supports the growth of some anaerobes better than standard media such as Mueller–Hinton broth and may be used as an alternative medium for MIC studies.17 Using this medium we were able to read MICs for 93% of strains examined. We adjusted inocula prepared in each of the media to 104 cfu/spot, using a MacFarland standard tube. Such an approach may mean that the relative proportions of spores and vegetative cells in inocula may differ, but, for practicality reasons, this is the best option available. Geometric mean MICs were uniformly higher (2.5- to 5.4-fold) using Schaedler's broth as opposed to Mueller–Hinton broth. This difference probably reflects the improved growth of strains prepared in Schaedler's broth. The issue of how to determine MICs for anaerobes is clearly contentious,13,18 and for example, end-point determination can be difficult during reading, particularly when hazy growth occurs at several different concentrations.
We have shown previously that the endemic C. difficile strain (PCR ribotype 1) sporulates significantly more than other strains when cultured in Schaedler's broth, although no significant differences were observed when strains were cultured in a human faecal emulsion.19 The results of the present study indicate a further potential virulence determinant for this C. difficile strain, namely reduced antibiotic (fluoroquinolone) susceptibility compared with other strains found in the nosocomial setting. This may increase the likelihood of selection of C. difficile PCR ribotype 1 when antibiotic pressure exists. We are currently investigating whether the epidemic C. difficile clone examined here is more resistant to other antibiotics compared with other strains.
The broad-spectrum activity of new fluoroquinolones offers the possibility of single agent oral therapy for the empirical treatment of sepsis where Gram-positive and/or Gram-negative pathogens are suspected. Enhanced pneumococcal activity in particular will open the possibility of single agent treatment of respiratory tract infection.8,10 However, new agents should not increase the susceptibility of elderly patients to C. difficile infection if they are to have a useful role in this age group. Early generation fluoroquinolones, such as ciprofloxacin, have rarely been implicated as causing C. difficile diarrhoea.20 In the few case reports that have been published it is notable that there are often other explanations for gut flora disturbance, such as salmonella gastroenteritis.21,22 Interestingly, ciprofloxacin has occasionally been used successfully to treat C. difficile infection.23 The likelihood of an antibiotic inducing C. difficile infection will be determined in part by the activity of the antibiotic against gut flora, but also its activity against C. difficile. The increased activity of new fluoroquinolones against anaerobes may therefore be both an advantage and a disadvantage. Superior anti-C. difficile activity must be balanced against the possible perturbation of the anaerobic gut flora. In this context, it should be noted that clindamycin-associated C. difficile diarrhoea has been shown recently to be associated with an epidemic clindamycin-resistant C. difficile strain.24
High faecal concentrations (>1000 µg/g) are obtained with the fluoroquinolones studied in the present study, with the exception of levofloxacin (up to 100 µg/g) given its predominant excretion via the urinary tract. However, questions remain concerning the absolute bioavailability of fluoroquinolones in the gut lumen given the effects of drug metabolism, low pH, anaerobiasis, bacterial inoculum size and antibiotic binding to organic matter. For example, it is known that faecal concentrations of ciprofloxacin may reach 2000–3000 µg/g, and yet while the MICs of most gut anaerobes do not exceed 64–128 mg/L, there is little or no reduction in the numbers of anaerobic flora following ciprofloxacin administration.25 Hence, the activity of fluoroquinolones on gut commensals or pathogens is probably much lower than expected from antibiotic concentration data alone.
This study was conceived before the market withdrawals on safety grounds of grepafloxacin and trovafloxacin. Pre-licensing trials have shown no evidence of an increased predisposition towards C. difficile infection in patients receiving moxifloxacin.26 It remains to be seen whether this observation, taken with the enhanced activity of moxifloxacin in comparison with other fluoroquinolones against C. difficile, means that this new generation fluoroquinolone is unlikely to be associated with C. difficile infection. One retrospective study, presented only in abstract form, documented an increased risk of C. difficile colitis associated with trovafloxacin use.27 It will therefore be important to monitor the use of new generation fluoroquinolones in elderly hospitalized patients where C. difficile is endemic. In particular, the reduced antibiotic susceptibility to all fluoroquinolones of the readily transmissible C. difficile strain documented here warrants further study.
|
Acknowledgments |
|---|
We thank Bayer for providing antibiotic powders and financial support for this study.
|
Notes |
|---|
* Correspondence address. Department of Microbiology, University of Leeds, Leeds LS2 9JT, UK. Tel: +44-113-233-5595; Fax: +44-113-233-5649; E-mail: markwi{at}pathology.leeds.ac.uk
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
1 . Wilcox, M. H. & Smyth, E. T. (1998). Incidence and impact of Clostridium difficile infection in the UK, 1993–1996. Journal of Hospital Infection 39, 181–7.
2 . Clostridium difficile in England and Wales: weeks 27–53/98. (1999). Communicable Disease Report CDR Weekly 9, 59.
3 . Wilcox, M. H., Cunniffe, J. G., Trundle, C. & Redpath, C. (1996). Financial burden of hospital-acquired Clostridum difficile infection. Journal of Hospital Infection 34, 23–30.[Web of Science][Medline]
4 . Riley, T. V. (1998). Clostridium difficile: a pathogen of the nineties. European Journal of Clinical Microbiology and Infectious Diseases 17, 137–41.[Web of Science][Medline]
5 . Anand, A., Bashey, B., Mir, T. & Glatt, A. E. (1994). Epidemiology, clinical manifestations, and outcome of Clostridium difficile-associated diarrhoea. American Journal of Gastroenterology 89, 519–23.[Web of Science][Medline]
6 . Settle, C. D., Wilcox, M. H., Fawley, W. N., Corrado, O. J. & Hawkey, P. M. (1998). Prospective study of the risk of Clostridium difficile diarrhoea in elderly patients following treatment with cefotaxime or piperacillin-tazobactam. Alimentary Pharmacology and Therapeutics 12, 1217–23.[Web of Science][Medline]
7 . Medicines Control Agency. (1994). Antibiotic-associated colitis. Current Problems in Pharmacovigilance 20, 7.
8 . Blondeau, J. M. (1999). Expanded activity and utility of the new fluoroquinolones: a review. Clinical Therapeutics 21, 3–40.[Web of Science][Medline]
9 . Edlund, C., Sabouri, S. & Nord, C. E. (1998). Comparative in vitro activity of BAY 12-8039 and five other antimicrobial agents against anaerobic bacteria. European Journal of Clinical Microbiology and Infectious Diseases 17, 193–5.[Web of Science][Medline]
10 . Woodcock, J. M., Andrews, J. M., Boswell, F. J., Brenwald, N. P. & Wise, R. (1997). In vitro activity of BAY 12-8039, a new fluoroquinolone. Antimicrobial Agents and Chemotherapy 41, 101–6.[Abstract]
11
.
Stubbs, S. L., Brazier, J. S., O'Neill, G. L. & Duerden, B. I. (1999). PCR targeted to the 16S–23S rRNA gene intergenic spacer region of Clostridium difficile and construction of a library consisting of 116 different PCR ribotypes. Journal of Clinical Microbiology 37, 461–3.
12 . Epidemic methicillin resistant Staphylococcus aureus. (1997). Communicable Disease Report Weekly 6, 191.
13 . National Committee for Clinical Laboratory Standards. (1993). Methods for Antimicrobial Susceptibility Testing of Anaerobic Bacteria —Third Edition: Approved Standard M11-A3. NCCLS, Villanova, PA.
14
.
Rafferty, M. E., Baltch, A. L., Smith, R. P., Bopp, L. H., Rheal, C., Tenover, F. C. et al. (1998). Comparison of restriction enzyme analysis, arbitrarily primed PCR, and protein profile analysis typing for epidemiologic investigation of an ongoing Clostridium difficile outbreak. Journal of Clinical Microbiology 36, 2957–63.
15 . Cohen, S. H., Tang, Y. J., Muenzer, J., Gumerlock, P. H. & Silva, J. (1997). Isolation of various genotypes of Clostridium difficile from patients and the environment in an oncology ward. Clinical Infectious Diseases 24, 889–93.[Web of Science][Medline]
16 . Nord, C. E. (1996). In vitro activity of quinolones and other antimicrobial agents against anaerobic bacteria. Clinical Infectious Diseases 23, Suppl. 1, S15–8.
17 . Summanen, P., Baron, E. J., Citron, D. M., Strong, C., Wexler, H. M. & Finegold, S. M. (1993). Susceptibility testing of anaerobic bacteria. In Wadsworth Anaerobic Bacteriology Manual, 5th edn, 111–27. Star Publishing Company, Belmont, CA.
18 . Finegold, S. M. (1997). Perspective on susceptibility testing of anaerobic bacteria. Clinical Infectious Diseases 25, Suppl. 2, S251–3.
19 . Fawley, W. N. & Wilcox, M. H. (1998). Study of the sporulation of a hospital endemic Clostridium difficile strain. Abstracts presented at the Fourth International Conference of the Hospital Infection Society, Edinburgh, 1998. Journal of Hospital Infection 40, Suppl. A, Abstract 3.6.1.1.
20
.
Golledge, C. L., Carson, C. F., O'Neill, G. L., Bowman, R. A. & Riley, T. V. (1992). Ciprofloxacin and Clostridium difficile-asociated diarrhoea. Journal of Antimicrobial Chemotherapy 30, 141–7.
21 . Bates, C. J., Wilcox, M. H., Spencer, R. C. & Harris, D. M. (1990). Ciprofloxacin and Clostridium difficile infection. Lancet 336, 1193.
22 . Hillman, R. J., Gopal Rao, G., Harris, J. R. W. & Taylor-Robinson, D. (1990). Ciprofloxacin as a cause of Clostridium difficile-asociated diarrhoea in an HIV-positive patient. Journal of Infection 21, 205–7.[Web of Science][Medline]
23 . Daniels, J. & Pristas, A. (1992). Successful treatment of Clostridium difficile colitis with ciprofloxacin. Journal of Clinical Gastroenterology 15, 176–7.[Web of Science][Medline]
24
.
Johnson, S., Samore, M. H., Farrow, K. A., Killgore, G. E., Tenover, F. C., Lyras, D. et al. (1999). Epidemics of diarrhea caused by a clindamycin-resistant strain of Clostridium difficile in four hospitals. New England Journal of Medicine 341, 1645–51.
25 . Nord, C. E. (1988). Effect of new quinolones on the human gastrointestinal microflora. Reviews of Infectious Diseases 10, Suppl. 1, S193–6.
26 . Springsklee, M., Reiter, C. & Mayer, J. M. (1999). Safety and tolerability profile of moxifloxacin. Clinical Microbiology and Infection 5, Suppl. 3, Abstract P0208.140.
27 . Hernandez, F., Khorana, A. A., Scott, J. D., Feuerstein, S., Forrest, A. & Freer, J. P. (1998). Incidence of C. difficile colitis in patients treated with antibiotics in a community hospital setting, with emphasis on trovafloxacin. Society for Healthcare Epidemiology of America Annual Meeting, San Francisco, 1998. Abstract.
Received 14 February 2000; returned 15 May 2000; revised 23 May 2000; accepted 26 June 2000
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Saxton, S. D. Baines, J. Freeman, R. O'Connor, and M. H. Wilcox Effects of Exposure of Clostridium difficile PCR Ribotypes 027 and 001 to Fluoroquinolones in a Human Gut Model Antimicrob. Agents Chemother., February 1, 2009; 53(2): 412 - 420. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. D. Baines, R. O'Connor, J. Freeman, W. N. Fawley, C. Harmanus, P. Mastrantonio, E. J. Kuijper, and M. H. Wilcox Emergence of reduced susceptibility to metronidazole in Clostridium difficile J. Antimicrob. Chemother., November 1, 2008; 62(5): 1046 - 1052. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Mutlu, A. J. Wroe, K. Sanchez-Hurtado, J. S. Brazier, and I. R. Poxton Molecular characterization and antimicrobial susceptibility patterns of Clostridium difficile strains isolated from hospitals in south-east Scotland J. Med. Microbiol., July 1, 2007; 56(7): 921 - 929. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Drudy, T. Quinn, R. O'Mahony, L. Kyne, P. O'Gaora, and S. Fanning High-level resistance to moxifloxacin and gatifloxacin associated with a novel mutation in gyrB in toxin-A-negative, toxin-B-positive Clostridium difficile J. Antimicrob. Chemother., December 1, 2006; 58(6): 1264 - 1267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. N. Fawley, P. Parnell, P. Verity, J. Freeman, and M. H. Wilcox Molecular Epidemiology of Endemic Clostridium difficile Infection and the Significance of Subtypes of the United Kingdom Epidemic Strain (PCR Ribotype 1) J. Clin. Microbiol., June 1, 2005; 43(6): 2685 - 2696. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.A. O'Connor, M. Kingston, M. O'Donovan, B. Cryan, C. Twomey, and D. O'Mahony Antibiotic prescribing policy and Clostridium difficile diarrhoea QJM, July 1, 2004; 97(7): 423 - 429. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Ackermann, A. Degner, S. H. Cohen, J. Silva Jr, and A. C. Rodloff Prevalence and association of macrolide-lincosamide- streptogramin B (MLSB) resistance with resistance to moxifloxacin in Clostridium difficile J. Antimicrob. Chemother., March 1, 2003; 51(3): 599 - 603. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Ortiz-de-Saracho, L. Pantoja, M. J. Romero, and R. Lopez Moxifloxacin-induced Clostridium difficile diarrhea Ann. Pharmacother., March 1, 2003; 37(3): 452 - 453. [Full Text] [PDF]
|
|
|
|
|
|
L. Dridi, J. Tankovic, B. Burghoffer, F. Barbut, and J.-C. Petit gyrA and gyrB Mutations Are Implicated in Cross-Resistance to Ciprofloxacin and Moxifloxacin in Clostridium difficile Antimicrob. Agents Chemother., November 1, 2002; 46(11): 3418 - 3421. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Mylonakis, E. T. Ryan, and S. B. Calderwood Clostridium difficile-Associated Diarrhea: A Review Arch Intern Med, February 26, 2001; 161(4): 525 - 533. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Freeman and M. H. Wilcox Antibiotic activity against genotypically distinct and indistinguishable Clostridium difficile isolates J. Antimicrob. Chemother., February 1, 2001; 47(2): 244 - 246. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||