Journal of Infection (2006) 53, 279e288

www.elsevierhealth.com/journals/jinf

Phenotypic and genotypic detection of ESBLmediated cephalosporin resistance in Klebsiellapneumoniae: Emergence of high resistanceagainst cefepime, the fourth generationcephalosporin

S.S. Grover a, Meenakshi Sharma a,*, D. Chattopadhya b,Hema Kapoor c, S.T. Pasha a, Gajendra Singh d

a Department of BiochemistryeBiotechnology, National Institute of CommunicableDiseases, 22-Shamnath Marg, New Delhi-110054, Indiab Department of Microbiology, National Institute of Communicable Diseases,22-Shamnath Marg, New Delhi-110054, Indiac Department of Microbiology, Safdarjung Hospital, New Delhi-110029, Indiad Department of Pharmaceutical Sciences, Guru Nanak Dev University, Amritsar,Punjab-143005, India

Accepted 5 December 2005Available online 20 February 2006

KEYWORDSKlebsiellapneumoniae;Cephalosporin;ESBL;PCR

Summary Objectives: Cephalosporins belonging to second and third generationare commonly used in India for the treatment of Klebsiella pneumoniae. Reporton resistance among K. pneumoniae strains to second and third generation cepha-losporins are on rise in this country, which has been attributed to emergence ofstrains expressing extended-spectrum b-lactamases (ESBLs). The aim of this studywas to evaluate the in vitro susceptibility of K. pneumoniae to broad-spectrumcephalosporins particularly to cefepime, a recently introduced fourth generationcephalosporin in relation to ESBL production.Methods: This study has been carried out in two phases among K. pneumoniaestrains isolated between October 2001 and September 2002 (phase I, before mar-keting of cefepime in India) and between August 2003 and July 2004 (phase II, aftermarketing of cefepime in India). Minimum Inhibitory Concentration (MIC) was

* Corresponding author. Indian Council of Medical Research, Ansari Nagar, New Delhi-110029, India. Tel.: þ91 11 26589544; fax: þ9111 26588381.

E-mail address: sharmam@icmr.delhi.nic.in (M. Sharma).

0163-4453/$30 ª 2006 The British Infection Society. Published by Elsevier Ltd. All rights reserved.doi:10.1016/j.jinf.2005.12.001

280 S.S. Grover et al.

determined by a commercial strip containing gradient of antimicrobials (Strip E-test). Detection for ESBL production was carried out by DDST, E-test ESBL and PCR.Results: Antimicrobial resistance profile of K. pneumoniae strains to five cephalos-porins as analyzed by WHONET 5 identified 15 different resistance profiles amongthe 108 phase I isolates, ranging from resistance to none (19.44%) to all the fivecephalosporin (8.33%) and eight different resistance profiles among the 99 phaseII isolates, ranging from resistance to none (9.1%) to all the five cephalosporins(36.4%). Among the 108 phase I isolates a total of 71 (65.72%) and out of 99 phaseII isolates, a total of 87 (88.0%) could be identified as ESBL producers. Among theisolates, regardless of the phase of the isolation, those characterized by productionof ESBL showed overall higher frequency of resistance to cephalosporins (range19.7e85.9% and 51.7e100% in phase I and phase II, respectively) compared to thosefor ESBL non-producers (range 0e13.5% and 0e25% in phase I and phase II, respec-tively). Ten randomly selected isolates from the most common resistance pheno-types probably belonged to a single strain as evident by MIC patterns, genotypiccharacterization and resistance profile to non-cephalosporin group of antimicro-bials thereby pointing out the possibility of an outbreak.Conclusions: PCR may be regarded as a reliable method for detection of ESBL sincein addition to the strains that could be identified as ESBL producers by DDST andE-test ESBL; PCR could demonstrate ESBL production among additional 32 strains(15 in phase I and 17 in phase II). Continued uses of cephalosporin group appear tobe a potential risk factor for emergence of ESBL producing K. pneumoniae strains.In addition, as noted in the present study, the rise of resistance to cefepime thathas been introduced recently in this country for therapeutic use could be of concern.ª 2006 The British Infection Society. Published by Elsevier Ltd. All rights reserved.

Introduction

In India cephalosporins belonging to second andthird generation are commonly used for the treat-ment of infections due to Klebsiella pneumoniae.1

However, resistance to these cephalosporinsamong K. pneumoniae strains is on the rise in thiscountry.2 This has been attributed to emergenceof strains producing extended-spectrum b-lacta-mases (ESBLs), the enzymes capable of hydrolyzingbroad-spectrum cephalosporins.3 ESBLs are plas-mid-mediated enzymes that have evolved throughpoint mutation of amino acids in parent TEM andSHV enzymes.4 The ESBL producing K. pneumoniaestrains are being reported from all over the worldincluding India.5e7 However, reports on monitoringof the trends in cephalosporin resistance and ESBLproduction among K. pneumoniae strains in identi-cal setting over a time period are limited and to thebest of our knowledge there is none from India.Furthermore, there is no study from India onthe pattern of resistance to the fourth generationcephalosporin i.e. cefepime, which has beenrecently introduced in India.8

This study was carried out to find out relativestatus of cephalosporin resistance and ESBL pro-duction in K. pneumoniae strains isolated duringtwo periods i.e. before (October 2001 to September2002) and after (August 2003 to July 2004)the introduction of cefepime in Indian market

employing the techniques viz. Minimum InhibitoryConcentration (MIC) by E-test, ESBL detection bydouble-disc synergy test (DDST) as well as byE-test and by Polymerase Chain Reaction (PCR) fordetection of blaTEM and blaSHV genes.

Materials and methods

A total of randomly selected 207 strains of K.pneumoniae isolated from Safdarjung hospital,New Delhi, comprising 108 strains isolated duringthe period October 2001 to September 2002 (theperiod before introduction of cefepime in Indianmarket, phase I) and 99 isolated from August2003 to July 2004 (the period after marketing cefe-pime in Indian market, phase II8), were confirmedby culture and biochemical characteristics.9 Theisolates were from clinical specimens/sites viz. pus,wound, pleural fluid, urine and tracheal aspirate.

Determination of minimum inhibitoryconcentration (MIC) and profile ofresistance

MIC was determined by commercial MIC E-test stripcontaining gradient of antimicrobial concentra-tions, 0.01e256 mg/ml (E-test Strip, AB BIODISK,Solna, Sweden), as per the procedure of manufac-tures.10 Briefly, the MIC E-test is based on diffusion

Phenotypic and genotypic detection of cephalosporin resistance in K. pneumoniae 281

method that enables MIC values to be estimateddirectly using single culture plate. The MIC E-teststrips were applied on MuellereHinton agar(MHA) plate inoculated with concentration ofw104 CFU/ml of K. pneumoniae strain (culture ad-justed to 0.5 McFarland standard). Escherichia coliATCC 25922 was used as a standard strain. The MICwas read at the point of intersection of the zone ofinhibition with the point on scale of the strip. Thefive antimicrobials selected for MIC E-test in thepresent study belonging to cephalosporin groupwere cefoxitin, cefuroxime, cefotaxime, ceftazi-dime and cefepime.

MIC was determined according to the guidelinesof the Clinical and Laboratory Standards Institute(CLSI; formerly known as the National Committeefor Clinical Laboratory Standards). The MIC50 andMIC90 values were defined as MICs at which 50and 90% of strains, respectively, were inhibited.The percentages of susceptible, resistant or inter-mediate strain were also calculated on the basis ofbreakpoints according to CLSI guidelines i.e. forceftazidime and cefotaxime strains showing MIC(mg/ml) �1 as susceptible, >1 to <2 as intermedi-ate and �2 as resistant; for cefuroxime strainsshowing MIC (mg/ml) �4 as susceptible, >4 to <32as intermediate and �32 as resistant; for cefoxitinand cefepime strains showing MIC (mg/ml), �8 assusceptible, >8 to <32 as intermediate and �32as resistant. The data on sensitivity profile wereanalyzed by WHONET 5 software.

Tests for ESBL production

Double-disc synergy test (DDST)DDST was employed by placing four discs viz.cefotaxime (30 mg), ceftazidime (30 mg), cefepime(30 mg) and aztreonam (30 mg) discs radially at adistance of 30 mm each from an augmentin disc(amoxicillin 20 mg plus 10 mg clavulanic acid) ona lawn culture of K. pneumoniae isolate on MHAplate. Cefotaxime, ceftazidime and aztreonamdiscs were also placed on a parallel MHA plate butwithout placing augmentin disc using the sameinoculum as employed in the plate with augmentindisc. Both the plates were incubated aerobically at37 �C. The strain was considered as ESBL producer,if the zone size around any of the discs increased by�5 mm towards augmentin disc compared to thatproduced in absence of augmentin.11,12

E-test ESBLThe production of ESBL was also detected by usingE-test ESBL strips (PDM Epsilmeter, AB Biodisk,Solna, Sweden) as described earlier.13 Briefly,

procedure for E-test ESBL involved application oftwo strips with low range cephalosporin (cefo-taxime) and high range cephalosporin (ceftazi-dime). The low range ESBL strip containedcefotaxime (CT) gradient at one end (0.25e16 mg/ml) and a gradient of cefotaxime (0.016e1 mg/ml) plus 4 mg/ml of clavulanic acid, an inhib-itor of ESBL, at the other end (CTL). The high rangeESBL strip had ceftazidime (TZ) gradient at oneend (0.5e32 mg/ml) and a gradient of ceftazidime(0.064e4 mg/ml) plus clavulanic acid (4 mg/ml) atother end (TZL). The strain was considered ESBLproducer if MIC of CT was recorded as �0.5 mg/ml and MIC ratio of CT/CTL as �8 or MIC of TZ as�1 mg/ml and MIC ratio of TZ/TZL as �8. Thestrains were considered as ESBL non-producers ifCT/CTL ratio was recorded as <8 or MIC of CT as<0.5 and ratio TZ/TZL as <8 or MIC of TZ< 1were found. K. Pneumoniae ATCC 700603 (ESBLpositive genotype) was used as a positive control.

Molecular detection and identification of ESBL(blaTEM and blaSHV genes) by polymerase chainreaction (PCR) amplificationPCR amplification of ESBL genes blaTEM and blaSHV

was carried out as described earlier.13,14 GenomicDNA was purified by phenol extraction and ethanolprecipitation method.15 The known sequences ofESBL genes were used for designing primers. Oligo-nucleotides primers used for PCR assay were asfollow:

Primer A: 50-ATA AAA TTC TTG AAG ACG AAA-30

Primer B: 50-GAC AGT TAC CAA TGC TTA ATC A-30

Primer C: 50-GGG TAA TTC TTA TTT GTC GC-30

Primer D: 50-TTA GCG TTG CCA GTG CTC-30

Primers A and B were specific for blaTEM whileprimers C and D were specific for blaSHV.

The 50 ml PCR reaction mixture contained 10 mMTriseHCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 1 U ofTaq DNA polymerase (AmpliTaq Gold, Applied Bio-systems), 0.2 mM deoxynucleoside triphosphate,0.2 mM concentration (each) of primers and 5 ml ofthe DNA sample. The PCR was carried out underthe following conditions: 35 cycles with heat dena-turation at 95 �C for 60 s, primer annealing at 58 �Cfor 60 s, and DNA extension at 72 �C for 90 s bya DNA thermal cycler (model 9600; Applied Biosys-tems, Foster). The amplified DNA was separatedby 2% agarose gel electrophoresis, stained withethidium bromide, and visualized by UV transillumi-nation. The 930 bp and 1079 bp bands were that ofSHV and TEM enzymes, respectively,14,16,17 (Fig. 1).

Random selections of 10 strains of K. pneumo-niae among two most common phenotypes of

282 S.S. Grover et al.

1000bp

500bp

M L1 L2 L3 L4 L5 L6 NC

TEM 1079bpSHV 930 bp

Figure 1 ESBL by PCR (TEM and SHV). M: marker 100 bp, NC: negative control, L1 and L3: SHV e 930 bp. L2, L4, L5,and L6: TEM e 1079 bp.

resistance of phase II (2003e2004) were analyzed interms of period of isolation, MIC profile to cephalo-sporins by E-test as well as resistance profile toa panel of non-cephalosporins group of antimicro-bials viz., ampicillin (10 mg), aztreonam (30 mg);gentamicin (10 mg), tobramycin (10 mg), ofloxacin(5 mg), trimethoprim (5 mg), tetracycline (30 mg),chloramphenicol and augmentin (amoxicillin 20 mgplus clavulanic acid 10 mg), by disc diffusion test.PCR products of all the 10 random selected strainswere subjected to sequencing of blaTEM genes.14

The amplicons were purified with PCR purificationKit (QIAGEN, Hidden, Germany) and directlysequenced with an ABI PRISM 310 sequencer, DyeTerminator cycle sequencing kit (Perkin Elmer),using specific blaTEM primers as per manufacturer’srecommendations. The sequences thus obtainedwere edited and aligned using Sequence NavigatorSoftware followed by blast at the National Centerfor Biotechnology Information website search(http://www.ncbi.nlm.nih.gov/blast).

Results

Frequency of resistance againstcephalosporins in K. pneumoniae strains

The frequency of resistance among K. pneumoniaeisolated during the two phases of the present studyfor the five cephalosporins viz. cefoxitin, cefurox-ime, cefotaxime, ceftazidime, and cefepime asdetected by MIC E-test are shown in Table 1. Itwas observed that incidence of resistance amongK. pneumoniae isolates for each of the five cepha-losporins increased significantly between the twostudy periods i.e. 2001e2002 and 2003e2004(Table 1). Resistance for recently introduced cefe-pime was lowest among the five cephalosporins

tested in phase I although there was a sharp in-crease in incidence of resistance (more than 4.5fold increase) among the K. pneumoniae strainsisolate in phase II as a result of which this antimi-crobial could no longer be recorded as least resis-tant among the various cephalosporins tested.MIC50, MIC90 and geometric mean of MIC amongthe phase II isolates for cefepime were higher ascompared to phase I isolates.

During the two phases of the study the overallcephalosporin resistance increased from 80.56(phase I) to 90.91% (phase II). Cephalosporin re-sistance profile of five cephalosporins as analyzedby WHONET 5 for phase I isolates identified 15different resistance profiles ranging from resis-tance to none (21 out of 108, i.e. 19.4%) and to one(15 out of 108 strains i.e. 13.8%) and to all the fivecephalosporins (9 out of 108 i.e. 8.3%; Table 2). Onthe other hand for phase II isolates a limited rangeof co-resistance profile (8 different profiles) couldbe recorded ranging from resistance to none (9 outof 99 i.e. 9.1%), to one (3 out of 99 i.e. 3.0%) and toall five (36 out of 99 i.e. 36.4%).

It was observed that co-resistance pattern ofisolates for all five cephalosporins was consider-ably higher among the phase II isolates (36 out of90 resistant strains i.e. 40.0%) compared to phase Iisolates (9 out of 87 resistant strains i.e. 10.3%)(c2¼ 20.47; p< 0.001). It was observed that boththe phase I and II resistances to cefepime werealways accompanied by co-resistance to one ormore of the remaining cephalosporins.

ESBL production by DDST, E-test ESBLand correlation with PCR

Out of 108 isolates of K. pneumoniae strains fromphase I, a total of 56 (51.8%) and out of 99 from

Phenotypic and genotypic detection of cephalosporin resistance in K. pneumoniae 283

Table

1Anti

mic

robia

lsu

scepti

bilit

ypatt

ern

of

K.

pneum

onia

efo

ra

spect

rum

of

five

cephalo

spori

ns

Anti

mic

robia

lage

nt

Bre

akp

oin

tsYear

of

isola

tion

Susc

epti

ble

no.

(%)

Resi

stant

no.

(%)

Inte

rmedia

teno.

(%)

MIC

50

MIC

90

Geom

etr

icm

ean

SI

R

Cefo

xiti

n(s

eco

nd

genera

tion)

�8

>8

to32

�32

2001

e20

0277

(73.

1)21

(19.

4)10

(7.5

)8

326.

601

2003

e20

0451

(51.

5)48

(44.

2)0

(0)

851

223

.68

Cefu

roxi

me

(seco

nd

genera

tion)

�4

>4

to32

�32

2001

e20

0234

(31.

5)53

(49.

1)21

(19.

4)16

256

18.9

8820

03e

2004

15(1

5.2)

84(8

4.2)

0(0

)51

251

215

7.11

Cefo

taxi

me

(thir

dge

nera

tion)

�1

>1

to2

�2

2001

e20

0246

(42.

6)62

(57.

4)0

(0)

432

3.23

120

03e

2004

12(1

2.1)

87(8

7.9)

0(0

)32

256

44.9

8Ceft

azi

dim

e(t

hir

dge

nera

tion)

�1

>1

to2

�2

2001

e20

0235

(32.

4)73

(67.

6)0

(0)

612

86.

525

2003

e20

0412

(12.

1)87

(87.

9)0

(0)

6425

655

.98

Cefe

pim

e(f

ourt

hge

nera

tion)

�8

>8

to32

�32

2001

e20

0285

(78.

7)15

(13.

9)8

(7.2

)4

322.

032

2003

e20

0433

(33.

3)66

(66.

7)0

(0)

6425

615

.48

S¼

sensi

tive

,I¼

inte

rmedia

te,

and

R¼

resi

stance

.

phase II a total of 70 (70.7%) could be identified asESBL producers by DDST and E-test ESBL. The PCRdemonstrated blaTEM, blaSHV or both in all thestrains detected to be ESBL producers by DDST andE-test ESBL (56 strains in phase I and 70 in phaseII). In addition 15 and 17 strains in phase I and IIwere also confirmed either having blaTEM, blaSHV

or both the genes (Fig. 1). Thus taking all the testsfor ESBL detection into account a total of 71(65.7%) and 88 (88.8%) of the K. pneumoniae strainin phase I and phase II, respectively, could be iden-tified as ESBL producers, thereby showing frequen-cy of ESBL producing strains to be higher in phase IIcompared to phase I (c2¼ 18.37; p< 0.01).

Distribution of resistance for cephalosporinsamong the ESBL producers and ESBLnon-producers

It was observed that among the phase I and phase IIisolates of K. pneumoniae, ESBL producers showedhigher frequency of resistance as compared to ESBLnon-producers regardless of the category of cepha-losporins tested (Table 3). Moreover, the degree ofresistance, as assessed in terms of distribution ofMICs showed that ESBL producing K. pneumoniaestrains had resistance extended up to much higherconcentration of cephalosporins compared toESBL non-producers in both phase I and phase II(Tables 4 and 5).

It was observed that 10 randomly selectedstrains from the most common phenotypes profilewere isolated within a period of 3 months duringthe second phase of study. These strains hadsimilar distribution of MIC to the cephalosporinand non-cephalosporin groups of antimicrobialsincluded in the present study (Table 6). All the10 amplicons were sequenced, analyzed for thededuced amino acid sequences and blast at web-site search (http://www.ncbi.nlm.nih.gov/blast)confirmed ESBL variant TEM-104.

Discussion

With the introduction of each new class of b-lactam antimicrobial, new b-lactamases haveemerged leading to resistance to the correspond-ing class of antimicrobial.18 For example, wide-spread use of cefotaxime and ceftazidime aftertheir introduction in the 1980s was followed byquick emergence of resistance to these twoagents19 in the causative organisms. ESBLs of TEMand SHV type20,21 have been implicated for theseresistance patterns.

284 S.S. Grover et al.

Table 2 Resistance profile of five cephalosporins

Co-resistance profile Strains (2001e2002), n¼ 108 Strains (2003e2004), n¼ 99

Number ofcephalosporins

Cephalosporins Number % Number %

NIL NIL 21 (19.4) 9 (9.1)1 CAZ 5 (4.6) e e1 CXA 10 (9.2) 3 (3.0)2 FOX, FEP 1 (0.93) e e2 CAZ, FEP 1 (0.93) e e2 CTX, CAZ 6 (5.6) e e2 CXA, CAZ 4 (3.7) e e2 CXA, CTX 3 (2.8) e e3 FOX, CXA, CAZ 2 (1.9) e e3 CXA, CAZ, FEP 1 (0.93) e e3 CTX, CAZ, FEP e e 3 (3.0)3 CXA, CTX, CAZ 18 (16.7) 12 (12.2)4 FOX, CXA, CAZ, FEP 1 (0.9) e e4 FOX, CTX, CAZ, FEP e e 3 (3.0)4 FOX, CXA, CTX, CAZ 16 (14.8) 9 (9.1)4 CXA, CTX, CAZ, FEP 10 (9.2) 24 (24.2)5 FOX, CXA, CTX, CAZ, FEP 9 (8.3) 36 (36.4)

FOX e cefoxitin, CXA e cefuroxime, CTX e cefotaxime, CAZ e ceftazidime, and FEP e cefepime.

Prior to 2002, globally the reported range ofresistance in cross-sectional studies was between27 and 60% for cefoxitin, 42 and 55% for cefurox-ime, 26 and 49% for cefotaxime, 40 and 57% forceftazidime and 14 and 20% for cefepime.22-25 Theprevalence of resistance among phase I isolates(2001e2002) of our study to cephalosporins is inaccordance with these reports. However, limiteddata are available to record the trends in cephalo-sporin resistance over a period of time, especiallyin developing countries like India. In the presentstudy, resistance to all cephalosporins increasedconsiderably between the two phases of the study,i.e. from 2001e2002 to 2003e2004. These findingsare in consistence with those from United States,where rate of ceftazidime resistance increasedfrom 1.5% in 1987 to 3.6% in 1991 and by 1993 asmany as 20.0% of strains were resistant to thiscephalosporin.26 Similarly the resistance rate ofcefotaxime increased from 18.0 to 43.0% between1994 and 2001 in China.25 Consequently, the fourthgeneration cephalosporins like cefepime, were

developed and introduced in therapy in 1993 andhad broad-spectrum activity against K. pneumo-niae.27 Cefepime is known to possess several ad-vantages over the third generation ones viz. highintrinsic potency due to rapid penetration intoperiplasmic space, broader spectrum of activityagainst multi-resistant gram-negative bacteria in-cluding Klebsiella species and low potential forb-lactamase induction.27 This antimicrobial agentwas not marketed in India till August 20038 andto the best of our knowledge there has been no re-port regarding its effectiveness in Indian setting.Resistance rates for cefepime among ESBL produc-ing K. pneumoniae in various geographic regionsshow a wide range from 5.6% in Canada to 50.4%Latin America.28,29

In the present study degree of resistance tocefepime was the lowest (13.9%) among thecephalosporins tested before its introduction intothe Indian market. However, the alarming increaseof cefepime resistance to 67.7% either singly or incombination with one or other cephalosporins

Table 3 Percentage of resistance to different cephalosporins

Phase Cefoxitin Cefuroxime Cefotaxime Ceftazidime Cefepime

I ESBL producer 29.60 69.00 85.90 100.0 19.70ESBL non-producer 0.00 13.50 8.10 5.40 8.10

II ESBL producer 51.70 93.10 100.00 100.00 58.60ESBL non-producer 0.00 25.00 0.00 0.00 0.00

Phenotypic and genotypic detection of cephalosporin resistance in K. pneumoniae 285

Table 4 Distribution of MIC of cephalosporins of K. pneumoniae strains isolates in phase I (2001e2002) relation toESBL production

And represent the ESBL producer and ESBL non-producers, respectively, in this table. Shaded area represents MIC of resistantstrains.

following the introduction of drug in the Indianmarket is a cause of concern. This could be due toconsiderable increase in prevalence of ESBL pro-ducing K. pneumoniae isolates encountered by us,since cefepime resistance has been reported to beassociated with presence of ESBL.25

The frequency of ESBL producing strains ofK. pneumoniae reported from various parts of theworld during 1996e2004 ranged from 10 to55%30,31 which is relatively lower compared to thefrequency encountered in the present study(76.8% from 2001 to 2004). The frequency of ESBLproducing K. pneumoniae strains in Lebanon in-creased from 7.5 to 24.0% between 1997 and2001,32 in China from 11.0 to 34.0% between 1994and 200125 and in Thailand from 37.0 to 50.0% be-tween 1994 and 199933 while in India rates ofESBL production among K. pneumoniae increasedfrom 5.3 to 61.0% between 1997 and 2004.1,34,35

One study from India, although reported ESBL pro-duction as high as 92.5% among K. pneumoniaestrains the number of isolates tested was only27.36 In our study, an increase by a margin of23.1% in ESBL production was observed (from65.7% in 2001e2002 to 88.8% in 2003e2004).

Furthermore, resistance rates for cephalospor-ins in this study were much higher between ESBLproducers compared to ESBL non-producers.

Similarly, in a study carried out in Spain in 1998, itwas observed that 45.0 and 86.0% of cefotaxime andceftazidime resistant K. pneumoniae strains, re-spectively, were also ESBL producers while all ESBLnon-producers were sensitive to these two antimi-crobial.37,38 Our study for the first time describes100% resistance to cefotaxime and ceftazidime inESBL resistant strains. The higher resistance ratesto all five cephalosporins in ESBL producing strainsin phase II of this study underlie the limited scopeof these drugs in treatment of K. pneumoniae infec-tions in India.

In this study, we observed that DDST and E-testESBL had similar sensitivities although the formerbeing the relatively cheaper method can be usedeffectively in hospital setting. PCR has beenconsidered to be a reliable method for detectionof ESBL compared to DDST and E-test ESBL13,14e16

which is in accordance with the observation inthe present study since, in addition to the strainsthat could be identified as ESBL producers byDDST and E-test ESBL, PCR could identify ESBL pro-duction in the additional 32 strains (15 in phase Iand 17 in phase II). A recent study by Stewardet al. showed that out of 139 K. pneumoniaeisolates from hospital 138 (99.3%) demonstratedeither blaTEM, blaSHV or both by PCR compared to117 (84.2%) by DDST and 116 (83.50%) by broth

286 S.S. Grover et al.

Table 5 Distribution of MIC of cephalosporins of K. pneumoniae strains isolates in phase II (2003e2004) relationto ESBL production

And represent the ESBL producer and ESBL non-producers, respectively, in this table. Shaded area represents MIC of resistantstrains.

microdilution tests with cefepime and cefepimeplus clavulanic acid, respectively, suggesting,thereby PCR to be an excellent screening toolfor studying ESBL containing strains of K.pneumoniae.12

It appeared that the most common resistanceprofile observed in the present study was linkedwith the possibility of outbreak due to a singlecirculating strain as evident by relatively short

duration within which they were isolated as well asby the similar pattern of MIC. In case of all these 10strains collected randomly MIC values for CXA,CTX, CAZ, FEP were recorded as >256, >16, >32,and >256 mg/ml, respectively. But for cefoxitinMIC value of eight resistant strains was >32 mg/ml and two strains showed relative lack of resis-tance to cefoxitin as MIC 8 mg/ml which may beconsidered as biological resistant to cefoxitin.39

Table 6 Random selected 10 strains of phase II (2003e2004)

Patient ID # Co-resistance profile MIC mg/ml (cephalosporins) blaTEM gene

Cephalosporins Non-cephalosporins FOX CXA CTX CAZ FEP

2004-04-13,2004-04-23

CXA, CTX,CAZ, FEP

AMP, ATM, GEN, TOB,OFX, TMP, TCY, AMC,CHL

8 >256 >16 >32 >256 TEM-104,TEM-104

2004-04-29,2004-04-33,2004-05-40,2004-05-42,2003-05-44,2004-06-47,2004-06-52,2004-06-54

FOX, CXA,CTX, CAZ, FEP

AMP, ATM, GEN, TOB,OFX, TMP, TCY, AMC,CHL

>32 >256 >16 >32 >256 TEM-104TEM-104TEM-104TEM-104TEM-104TEM-104TEM-104TEM-104

FOX e cefoxitin, CXA e cefuroxime, CTX e cefotaxime, CAZ e ceftazidime, FEP e cefepime, AMP e ampicillin, ATM e aztreonam,GEN e gentamicin, TOB e tobramycin, OFX e ofloxacin, TMP e trimethoprim, TCY e tetracycline, CHL e chloramphenicol, andAMC e augmentin.

Phenotypic and genotypic detection of cephalosporin resistance in K. pneumoniae 287

It has been demonstrated that a strain showingresistance to cefepime mostly shows resistanceto all the remaining cephalosporins.29 Their profileof resistance to non-cephalosporin antimicrobialsand harboring of TEM-104 gene as evident by ge-netic analysis further strengthened the possibilityof outbreak due to single circulating strain.

In conclusion, we feel that uncontrolled use ofcephalosporin group appears to be a potential riskfactor for emergence of ESBL producing strains ofK. pneumoniae is a matter of concern. Several stud-ies have shown that reducing use of ceftazidime orcephalosporins decreases the frequency of ESBLproducing K. pneumoniae strains.40e43 The univer-sal presence of ceftazidime resistance in all cepha-losporin resistant strains of K. pneumoniae in ourstudy appears to be hallmark of Indian strains. Inaddition it is also felt that cefepime, the fourthgeneration cephalosporin that has been newlyintroduced for therapy in this country, needs tobe used with caution in treating K. pneumoniaedue to the resistance encountered against this anti-microbial agent and preventing of a possible out-break of a highly resistant K. pneumoniae strain.

References

1. Ananthan SA, Alavandi SV. Extended-spectrum b-lactamaseproduction and multi-drug resistance in Klebsiella speciesisolated from children under five with intestinal and extraintestinal infections. Indian J Med Res 2001;113:181e5.

2. Abigail S, Mathai E, Jasdasan MV, John TJ. Ceftazidimeresistance among Klebsiella pneumoniae in South India.Indian J Med Res 1995;102:53e5.

3. Rice LB, Bonomo RA. b-Lactamases: which ones are clinicallyimportant? Drug Resist Updat 2001;7:333e6.

4. Nathisuwan S, Burgess DS, Lewis II JS. ESbL: epidemiology,detection and treatment. Pharmacotherapy 2001;21:920e8.

5. Jacoby GA, Medeiors AA. More extended-spectrum b-lacta-mases. Antimicrob Agents Chemother 1991;35:1697e704.

6. Chanawong A, Fatima FM, Zali FM, Heritage J, Lulitanond A,Hawkey PM. SHV-12, SHV-5, SHV-2a and VEB-1 extended-spectrum b-lactamases in gram-negative bacteria isolatedin a university hospital in Thailand. J Antimicrob Chemo-ther 2001;48:839e52.

7. Chang FY, Siu LK, Fung CP, Huang MH, Monto HO. Diversityof SHV and TEM b-lactamases in Klebsiella pneumoniae:gene evolution in Northern Taiwan and two novel b-lacta-mases, SHV-25 and SHV-26. Antimicrob Agents Chemother2001;45:2407e13.

8. MIMS, India. In: Gulati CM, editor. Monthly index of medicalspecialties. New Delhi: Indraprashta Press; 2003.

9. Farmer III JJ. Enterobacteriaceae: introduction and identifi-cation. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC,Yolken RH, editors. Manual of clinical microbiology. 6th ed.Washington D.C.: American Society for Microbiology; 1995.p. 438e44.

10. Roseblatt JE, Gustafson DR. Evaluation of the E-test for sus-ceptibility testing of anaerobic bacteria. Diagon MicrbiolInfect Dis 1995;22:279e84.

11. Jarlier V, Nicolas MH, Fournier G, Philippon A. ESBLs confer-ring transferable resistance to newer b-lactam agents inEnterobacteriaceae: hospital prevalence and susceptibilitypatterns. Rev Infect Dis 1988;10:867e78.

12. Steward CD, Rasheed JK, Susannah KU, Biddle JW,Raney PM, Anderson GJ, et al. Characterization of clinicalisolates of Klebsiella pneumoniae from 19 laboratoriesusing the National Committee for Clinical Laboratory Stand-ards extended spectrum b-lactamases detection methods.J Clin Microbiol 2001;39:2864e72.

13. Nuesh-Inderbinen MT, Hachler FHKH. Detection of genescoding for extended-spectrum SHV b-lactamases in clinicalisolates by molecular genetic method and comparison withE-test. Eur J Clin Microbiol Infect Dis 1996;15:398e402.

14. Mabilat C, Goussard S. Diagnostic molecular microbiology:principle and application. In: Persing DH, Smith TF,Tenover FC, White TJ, editors. PCR detection and identifi-cation of genes for extended-spectrum b-lactamases.Washington, DC: American Society for Microbiology; 1993.p. 553e63.

15. Sambrook J, Russell DW. Molecular cloning: a laboratorymanual. 3rd ed. In: Preparation and analysis of eukaryoticgenomic DNA, Vol. I Cold Spring Harbor, New York: ColdSpring Harbor Laboratory; 2001. p. 6.4e11.

16. Rasheed JK, Jay C, Matchock F, Berkowitz L, Weigel J,Crellin C, et al. Evolution of extended spectrum b-lactamresistance (SHV-8) in a strain of Escherichia coli during mul-tiple episodes of bacteremia. Antimicrob Agents Chemo-ther 1997;41:647e53.

17. Coque TM, Oliver A, Perez-Diaz JC, Baquero F, Canton R.Genes encoding TEM-4, SHV-2 and CTX-M-10 extended spec-trum b-lactamases are carried by multiple Klebsiella pneu-moniae clone in a single hospital (Madrid, 1989 to 2000).Antimicrob Agents Chemother 2002;46:500e10.

18. Bradford PA. Extended-spectrum b-lactamases in the 21stcentury: characterization, epidemiology and detection ofimportant resistance threat. Clin Microbiol Rev 2001;14:933e51.

19. Babini G, Livermore DM. Antimicrobial resistance amongstKlebsiella spp. from intensive care units in southern andWestern Europe in 1997e1998. J Antimicrob Chemother2000;45:183e9.

20. Sabate M, Miro E, Navarro F, Verges C, Aliaga R, Mirelis B,et al. b-Lactamases involved in resistance to broad-spec-trum cephalosporins in Escherichia coli and Klebsiella spp.Clinical isolates collected between 1994 and 1996, in Barce-lona (Spain). J Antimicrob Chemoth 2002;49:980e97.

21. Nukaga M, Kayako A, Hujer AM, Bonomo RA, Knox JR. Ultrahigh resolution structure of a class A b-lactamase: on themechanism and specificity of the extended-spectrumSHV-2 enzyme. J Mol Biol 2003;328:289e301.

22. Li J, Ma Y, Yu X. Klebsiella pneumoniae: epidemiology anal-ysis of risk factors for infections caused by resistant strains.Chin Med J 2002;115:1158e62.

23. Quale JM, Landman D, Bradford PA, Visalli M, Ravishankar J,Flores C, et al. Molecular epidemiology of citywide out-break of extended-spectrum b-lactamases in Klebsiellapneumoniae infection. Clin Infect Dis 2002;35:834e41.

24. Cao W, Tong MH, Wang JG. Extended-spectrum b-lacta-mases detection in enterobacteriaceae and antibiotic sus-ceptibility analysis. Hunan Yi Ke Da Xue Xue Bao 2002;27:77e8.

25. Chen MJ, Wang H. China nosocomial pathogens resistancesurveillance study group. Continuous surveillance of antimi-crobial resistance among nosocomial gram-negative bacillifrom intensive care units in China. Zhonghua Yi Xue ZaZhi 2003;85:375e81.

288 S.S. Grover et al.

26. Itokazu G, Ouinn J, Bell-Dixon C, Kahan F, Weinstein R.Antimicrobial resistance rates among aerobic gram-negative bacilli recovered from patients in intensive careunits. Evaluation of national post marketing surveillanceprogramme. Clin Infect Dis 1996;23:779e84.

27. Sanz MA, Lopez J, Lahuerta JJ, for the Spanish PETHEMAGroup. Cefepime plus amikacin versus piperacillinetazo-bactam plus amikacin for initial antibiotic therapy in hema-tology patients with febrile neutropenia: results of an open,randomized, multi-centre trial. J Antimicrob Chemother2002;50:79e88.

28. Winokur PL, Canton R, Clasellas JM, Legakis N. Variation inthe prevalence of strain expressing an extended-spectrumb-lactamases phenotype and characterization of isolatesfrom Europe, the Americas, and the western pacific region.Clin Infect Dis 2001;32:S94e103.

29. Yu WL, Praller MA, Winokur PL, Jones RN. Cefepime MIC asa predictor of an extended-spectrum b-lactamases type inKlebsiella pneumoniae Taiwan. Emerg Infect Dis 2002;8:522e4.

30. Rupp ME, Fey PD. Extended-spectrum b-lactamase (ESBL)producing Enterobacteriaceae; consider for diagnosis, pre-vention and drug treatment. Drugs 2003;63:353e65.

31. Casellas JM, Tome G, Bantar C, Bertolini P, Blazquez N,Borda N, et al. Argentinean collaborative multi-centerstudy on the in vitro comparative activity of piperacillinetazobactam against selected bacterial isolates recoveredfrom hospitalized patients. Diagn Microbiol Infet Dis 2003;47:527e37.

32. Daoud Z, Hakime N. Prevalence and susceptibility patterns ofextended-spectrum b-lactamases producing Escherichia coliand Klebsiella pneumoniae in a general university hospital inBeirut, Lebanon. Rev Esp Quimioiterap 2003;16:233e8.

33. Lulitanod A, Kaewkes W. Prevalence of extended-spectrumb-lactamases (ESBLs) production in a gram-negative bacillusisolated from Srinagarind hospital, Thailand. J Infect DisAntimicrob Agents 1999;16:115e9.

34. Shukla I, Tiwari R, Agarawal M. Prevalence of extended-spectrum b-lactamases producing Klebsiella pneumoniaein a tertiary care hospital. Indian J Med Microbiol 2004;22:87e91.

35. Mathai D, Rhomberg PR, Biedenbach DJ, Jones RN. IndianAntimicrobial Resistance Study Group. Evaluation of in vitroactivity of six broad-spectrum b-lactam antimicrobialagents tested against recent clinical isolates from India:a survey of ten medical center laboratories. Diagon Micro-biol Infect Dis 2002;44:367e77.

36. Manchanda V, Singh NP. Occurrence of AmpC b-lactometersamong gram-negative clinical isolates using a modifiedthree-dimensional test at Guru Tegh Bahadur Hospital,Delhi, India. J Antimicrob Chemother 2003;51:415e8.

37. Kim YK, Pai H, Lee HJ, Park SE, Choi EH, Kim J, et al. Blood-stream infection by extended-spectrum b-lactamasesproducing Escherichia coli and Klebsiella pneumoniae inchildren: epidemiology and clinical outcome. AntimicroAgents Chemother 2002;46:1481e91.

38. Pena C, Ardanuy C, Ricart A, Pallares R. Epidemiology andsuccessful control of a large outbreak due to Klebsiellapneumoniae producing extended-spectrum b-lactamases.Antimicro Agents Chemother 1998;42:53e8.

39. Pagani L, Perilli M, Migliavacca R, Luzzaro F, Amicosante G.Extended-spectrum TEM- and SHV-type b-lactamases-producing Klebsiella pneumoniae strains causing outbreaksin intensive care units in Italy. Eur J Microbol Infect Dis2000;19:765e72.

40. Rahal JJ, Urban C, Horn D, Freeman K, Segal-Maurer S,Maurer J, et al. Class restriction of cephalosporin resistancein Klebsiella. JAMA 1998;280:1233e7.

41. Rice LB, Willey SH, Papanicolaou GA, Medeiros AA,Eliopoulos GM, Moellering Jr RC, et al. Outbreak of ceftazi-dime resistance caused by extended-spectrum b-lacta-mases at Massachusetts chronic-care facility. AntimicrobAgents Chemother 1990;34:2193e9.

42. Naumovski L, Quinn JP, Miyashiro D, Patel M, Bush K,Singer SB, et al. Outbreak of ceftazidime resistance dueto a novel extended-spectrum b-lactamases in isolatesfrom cancer patients. Antimicrob Agents Chemother 1992;36:1991e6.

43. Pena C, Pujol M, Ricart A, Ardanuy C, Ayats J, Linares J,et al. Risk factors for faecal carriage of Klebsiella pneumo-niae producing extended-spectrum b-lactamases (ESBL-KP)in intensive care unit. J Hosp Infect 1997;35:9e16.