JOURNAL OF CLINICAL MICROBIOLOGY,0095-1137/99/$04.0010

Apr. 1999, p. 925–930 Vol. 37, No. 4

Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Serum Is More Suitable than Whole Blood for Diagnosisof Systemic Candidiasis by Nested PCR

M.-E. BOUGNOUX,1* C. DUPONT,1,2 J. MATEO,3 P. SAULNIER,4 V. FAIVRE,3

D. PAYEN,3 AND M.-H. NICOLAS-CHANOINE1

Departments of Microbiology1 and Internal Medicine,2 Hopital Ambroise-Pare, UniversiteParis V, 92100 Boulogne-Billancourt, Department of Anesthesiology and Intensive Care,

Hopital Lariboisiere, Universite Paris VII, 75010 Paris,3 and Department ofMicrobiology, Institut Gustave Roussy, 94805 Villejuif,4 France

Received 3 November 1998/Returned for modification 7 December 1998/Accepted 13 January 1999

PCR assays for the diagnosis of systemic candidiasis can be performed either on serum or on whole blood,but results obtained with the two kinds of samples have never been formally compared. Thus we designed anested PCR assay in which five specific inner pairs of primers were used to amplify specific targets on the rRNAgenes of Candida albicans, C. tropicalis, C. parapsilosis, C. krusei, and C. glabrata. In vitro, the lower limit of de-tection of each nested PCR assay was 1 fg of purified DNA from the corresponding Candida species. In rabbitswith candidemia of 120 minutes’ duration following intravenous (i.v.) injection of 108 CFU of C. albicans, thesensitivities of the PCR in serum and whole blood were not significantly different (93 versus 86%). In otherrabbits, injected with only 105 CFU of C. albicans, detection of candidemia by culture was possible for only 1min, whereas DNA could be detected by PCR in whole blood and in serum for 15 and 150 min, respectively. PCRwas more often positive in serum than in whole blood in 40 culture-negative samples (27 versus 7%; P < 0.05%).Lastly, experiments with rabbits injected i.v. with 20 or 200 mg of purified C. albicans DNA showed that PCRswere positive in serum from 30 to at least 120 min after injection, suggesting that the clearance of free DNAis slow. These results suggest that serum is the sample of choice, which should be used preferentially over wholeblood for the diagnosis of systemic candidiasis by PCR.

Systemic candidiasis is a major nosocomial infection inpatients given immunosuppressive chemotherapy for cancertreatment or organ transplantation and in patients undergoingheart or abdominal surgery (18). Patients with candidemiahave a poorer prognosis than those with nosocomial bactere-mia (19, 25). Mortality rates among those with systemic can-didiasis remain high, ranging from 50 to 80%, despite adequatetreatment (11, 26). In the absence of pathognomonic signs orsymptoms of systemic candidiasis, diagnosis is usually based onthe isolation of Candida species from blood cultures or tissuebiopsy specimens. However, since the sensitivity of blood cul-tures for diagnosis of systemic candidiasis is low at the earlystage of the infection, and since it has been shown that theprognosis is better when treatment is started early, it is usuallyrecommended that antifungal therapy be started as soon as astrong suspicion of systemic candidiasis exists (9, 16, 20). Onthe other hand, such empiric antifungal therapy may be un-necessarily toxic and costly, and it may increase the selectivepressure towards more-resistant Candida species (29). Thus,efforts have been made to develop more-sensitive methods forthe earliest possible diagnosis of systemic candidiasis. One ofthese involves the PCR method in which different targets ofCandida DNA have been tested: either single-copy genes suchas the actin (15), chitin synthase (14), HSP 90 (7), and lanos-terol-14 a-demethylase-encoding genes (3, 4) or multicopygenes such as the gene coding for rRNA (12, 13, 22, 23). Hy-bridization (8, 12, 22, 23) and nested PCR (4, 6) experimentshave been used to identify all the amplimers at the Candidaspecies level. The best of these assays are those which can

identify all the species most commonly involved in candidemia:Candida albicans, C. tropicalis, C. krusei, C. parapsilosis, andC. glabrata (8, 12–14, 22).

It has been demonstrated that PCR can be performed eitheron whole-blood samples (3, 4, 10) or on serum samples (5, 6,15). However, the efficiencies of the same PCR assay appliedsimultaneously to serum or whole blood have never been for-mally compared. These might not be equivalent, since theDNAs present in the two types of samples are probably differ-ent in origin. Indeed, only free template DNA should be de-tectable in serum samples, since fungal cells are eliminated bycentrifugation without having been lysed to release intracellu-lar DNA (6). By contrast, when whole-blood samples are used,both free DNA and intracellular DNA could be present whenthe sample is drawn from the patient. However, because of thepresence in blood of PCR inhibitors, such as hemoglobin, adecontamination step, including lysis of blood cells and wash-ing, is performed first. These steps probably eliminate freeCandida DNA, leaving intracellular Candida DNA as the solepossible target for the PCR assay (3, 4, 8, 23). Thus, dependingon the sample used, the origin of the detected DNA probablyvaries. This may result in a difference in the sensitivity of theassay and in its clinical significance. To our knowledge theseissues have not been fully investigated. This is why, in thepresent study, our efforts have been focused on that question.We have used a rabbit model of experimental candidemiaspecifically developed in this laboratory. We used DNA codingfor the 5.8S rRNA and the adjacent internal transcribed spacer(ITS) as the target for amplification, and we have comparedthe positivities of the PCR on whole-blood and serum speci-mens, using blood cultures as the reference assay.

(Part of this work was presented at the 37th InterscienceConference on Antimicrobial Agents and Chemotherapy, To-ronto, Ontario, Canada, 28 September to 1 October 1997 [2a]).

* Corresponding author. Mailing address: Service de Microbiologie,Hopital Ambroise-Pare, 9, avenue du General de Gaulle, 92100 Bou-logne-Billancourt, France. Phone: 33 (0) 1 49 09 55 45. Fax: 33 (0) 1 4909 59 21. E-mail: marie-elisabeth.bougnoux@apr.ap-hop-paris.fr.

925

on June 30, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

MATERIALS AND METHODS

Candida organisms. C. albicans ATCC 2091 was used for both in vitro and invivo experiments, whereas C. tropicalis ATCC 66029, C. glabrata ATCC 66032,C. krusei IP 208-52, C. parapsilosis IP 205-52, and clinical isolates of C. albicans(n 5 18), C. tropicalis (n 5 3), C. glabrata (n 5 6), C. parapsilosis (n 5 3), andC. krusei (n 5 2) were used for the in vitro experiments only.

Control DNA. DNAs from different species, including bacterial species (Pro-teus mirabilis, Enterobacter cloacae, Escherichia coli, Staphylococcus aureus, andMycobacterium tuberculosis), parasites (Toxoplasma gondii), and non-Candidafungal species (Aspergillus fumigatus, Cryptococcus neoformans, Pneumocystis ca-rinii, Trichophyton rubrum, and Microsporum canis), and human DNA preparedfrom amniotic fluid were used to determine the specificity of the primers de-signed to amplify Candida DNAs.

Preparation of Candida cells. Candida cells from stationary phase cultures inyeast extract-peptone-dextrose broth (18 h at 37°C, with shaking) were washedtwice in phosphate-buffered saline and counted in a hemocytometer. Countswere confirmed by agar plate counts.

DNA extractions. Candida DNA, extracted from broth culture as previouslydescribed (21), was stored at 280°C until use. Purified Candida DNA was quan-tified by using a GeneQuant RNA/DNA calculator (Pharmacia Biotech, Orsay,France).

DNA was extracted from whole blood as previously described (23), with slightmodifications. Briefly, 100 ml of whole blood was mixed with 100 ml of blood celllysis buffer (0.32 M sucrose, 10 mM Tris-HCl [pH 7.5], 5 mM MgCl2, and 1%Triton X-100) and centrifuged at 16,000 3 g for 5 min. The pellet was resus-pended in 200 ml of lysis buffer to which 7 ml of DNase I (10 mg/ml; BoehringerMannheim, Meylan, France) was added, in order to eliminate free DNA. Themixture was incubated for 1 h at 37°C, and the DNase was then inactivated byheating for 10 min at 85°C. After centrifugation, the pellet was resuspended in200 ml of TEG buffer (50 mM glucose, 25 mM Tris-HCl [pH 8], and 10 mMEDTA) containing 1.5 ml of lyticase (900 U/ml; Sigma, Saint Quentin Fallavier,France), and incubated for 1 h at 37°C. Three microliters of pronase E (15 mg/ml; Sigma) and 10 ml of 10% sodium dodecyl sulfate were added, and incubationwas continued for another hour at 37°C. DNA was then extracted with phenol-chloroform-isoamyl alcohol, precipitated with 2 volumes of ethanol, and dis-solved in 40 ml of sterile water.

DNA was extracted from serum as previously described (6), also with slightmodifications. Briefly, proteinase K (Sigma) and sodium dodecyl sulfate wereadded to 100 ml of serum at final concentrations of 15 mg/ml and 1%, respec-tively. The mixture was incubated for 1 h at 37°C and then boiled for 10 min toinactivate proteinase K. After phenol-chloroform-isoamyl alcohol extraction andethanol precipitation, DNA was dissolved in 40 ml of sterile water.

Oligonucleotide primers and PCR. The fungus-specific universal primers ITS1(59TCCGTAGGTGAACCTGCGG39) and ITS4 (59TCCTCCGCTTATTGATATGC 39) (27) were used as outer primers to amplify the intergenic transcribed

spacer regions of Candida species rRNAs. As indicated in Table 1, specific innerprimers were designed for C. albicans, C. parapsilosis, C. tropicalis, and C. kruseion the basis of the ITS1–ITS4 sequences derived from GenBank (respectiveaccession numbers: L47111, L47109, L47112, and L47113). For C. glabrata, innerprimers were designed from the ITS1–ITS3 sequence (GenBank accession no.L47110). The sequences of the specific inner primers used in the nested PCR,and the sizes of the amplification products, are indicated in Table 1.

PCR amplification was performed in a final volume of 25 ml by using a reactionmixture containing 50 mM KCl, 10 mM Tris-HCl (pH 8.3), 1.5 mM MgCl2, 100mM each deoxynucleoside triphosphate, and 1.25 U of Taq DNA polymerase(Boehringer Mannheim). For the first PCR, 10 pmol of each outer primer wasmixed either with 5 ml of DNA prepared from whole blood or serum or with 1ml of purified DNA. A 9600 thermal cycler (Perkin-Elmer, Saint Quentin-en-Yvelines, France) was used with the following temperature cycles: 95°C for 5min; then 30 cycles of 20 s at 95°C, 15 s at 55°C, and 65 s at 72°C; and a final cycleof PCR extension at 72°C for 5 min. For the nested PCR, 1 ml of the productobtained from the first amplification and 10 pmol of each inner primer was mixedin fresh reaction mixture. The second amplification was performed under theconditions described above, except for the annealing temperatures, which werespecific for each pair of inner primers, as indicated in Table 1. The nested PCRproducts were submitted to electrophoresis on a 1.5% agarose gel containingethidium bromide. Amplicon carryover was prevented by using aerosol-guardedpipette tips (ATGC Biotechnologie, Noisy Le Grand, France) and by carefullyseparating the DNA extraction area from the areas in which PCR reactionmixtures were prepared and the amplification and electrophoresis were per-formed. Appropriate negative controls, i.e., the DNA extraction and reactionmixture controls, were tested for each amplification reaction.

To avoid false interpretation of negative PCRs in rabbit blood samples, apositive internal control was designed. Two 39-mer composite primers containingM13mp18 phage sequences flanked at their 59 ends by C. albicans inner primers(Table 1) were synthesized (Genset, Paris, France). PCR was performed by usingthese composite primers on M13mp18 template DNA to generate an M13mp18fragment with a C. albicans sequence for each 59 position. When 90 pg of thisamplified product was added to the nested PCR reaction mixture, a 491-bpfragment was generated in the absence of inhibitors.

In vitro evaluation of the sensitivity and specificity of Candida nested PCR. Todetermine the detection limit for the purified DNAs of five Candida species, eachnested PCR was performed with 1 pg and with 100, 10, 1, and 0.1 fg of purifiedCandida DNA from each species tested. To check the inter-Candida speciesspecificity of the five nested PCRs, each nested PCR was performed with 100 ngof Candida DNA from the other four species. Candida species specificity wasthen evaluated by applying the five nested PCRs to 100 ng of control DNA.

To determine the smallest number of Candida cells for which Candida DNAwas detectable by nested PCR, different amounts of C. albicans and C. tropicalis

TABLE 1. Oligonucleotides used in nested PCR

Species (accession no.)and primers Sequence (59339) Nucleotide

positionaFragmentsize (bp)

Tempb

(°C)

C. albicans (L47111)Oligonucleotide 1: CAL1 AACTTGCTTTGGCGGTGGGC 73 386 66Oligonucleotide 2: CAL3 TGGACGTTACCGCCGCAAGC 439

C. tropicalis (L47112)Oligonucleotide 1: CTR1 ATTTCTTTGGTGGCGGGAGC 73 373 57Oligonucleotide 2: CTR3 GGCCACTAGCAAAATAAGCG 426

C. glabrata (L47110)Oligonucleotide 1: CGL1 ATGCTATTTCTCCTGCCTGC 128 234 57Oligonucleotide 2: CGL3 TGNATCCACTGGGAGAACTC 342

C. parapsilosis (L47109)Oligonucleotide 1: CPA1 GCCAGAGATTANACTCAACC 123 336 55Oligonucleotide 2: CPA3 GGAAGAAGTTTTGGAGTTTG 439

C. krusei (L47113)Oligonucleotide 1: CKR2 ACTACACTGCGTGAGCGGAA 43 360 55Oligonucleotide 2: CKR3 AAAAAGTCTAGTTCGCTCGG 383

Internal controlOligonucleotide 1: M1 AACTTGCTTTGGCGGTGGGCCCTCGGTTTCCTTCTGGTA 491 66Oligonucleotide 2: M3 TGGACGTTACCGCCGCAAGCGCGAACCTCCCGACTTGCG

a In the ITS sequence.b Annealing temperature.

926 BOUGNOUX ET AL. J. CLIN. MICROBIOL.

on June 30, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

cells were artificially inoculated into human and rabbit blood samples so thateach sample contained 104 to 1 CFU/ml.

Rabbit model. Fifteen male New Zealand rabbits weighing 2 to 2.5 kg werehoused in individual cages and anesthetized by injection of 15 mg of sodiumpentobarbital/kg of body weight into the marginal ear vein. Tracheotomy wasperformed, and the lungs were mechanically ventilated. Anesthesia and muscleparalysis were maintained by intermittent intravenous injection of 12.5 mg ofpentobarbital and 0.2 mg of pancuronium. Throughout the study period, rabbitsreceived 8 ml of 0.9% sodium chloride solution/h and 2 ml of 8.4% sodiumbicarbonate solution/h, by continuous intravenous infusion. A 14-gauge cannulawas inserted into the right carotid artery to measure mean arterial pressure, anda jugular vein catheter was inserted under sterile conditions for volume resusci-tation and blood sampling for PCR and cultures.

Experimental protocol. Rabbits were injected in the marginal ear vein eitherwith a 1-ml bolus of 108 (five rabbits) or 105 (six rabbits) C. albicans cells or with200 (two rabbits) or 20 (two rabbits) mg of purified C. albicans DNA. Beforeinjection and 1, 5, 15, 30, 60, 90, 120, 150, and 180 min thereafter, blood sampleswere collected in both EDTA-coated and dry tubes. One milliliter of the bloodcollected in the EDTA-coated tube was cultured on Sabouraud-chloramphenicolagar plates in order to count Candida cells. Another milliliter of the bloodcollected in the EDTA-coated tube and 500 ml of the serum obtained from thedry tube were used for independent DNA extractions and PCR assays. All PCRsand cultures were performed in duplicate.

Sequencing of the C. albicans fragment amplified by nested PCR. The frag-ment generated by inner primers of C. albicans from either purified DNA ofC. albicans ATCC 2091 or infected rabbit sera was sequenced from both endswith primers CAL1 and CAL3 (Genome Express, Lyon, France). These se-quences were compared to the C. albicans sequence available from GenBank(accession no. L47111).

Statistical analysis. The differences between PCR positivity rates on serumand on whole blood were tested by using the chi-square test with Yates’ correc-tion at the 5% level of significance.

RESULTS

In vitro sensitivity and specificity of Candida nested PCR.As indicated in Fig. 1, the limits of purified Candida DNA de-tection by nested PCR ranged from 1 to 0.1 fg, depending onthe Candida species. By inoculation of human and rabbit bloodsamples with either C. albicans or C. tropicalis cells, a PCR as-say was able to detect the Candida DNA extracted from as fewas 100 CFU/ml for C. albicans and 30 CFU/ml for C. tropicalis.There was no difference between the results of experimentsperformed with human and rabbit blood (data not shown).

The observed specificities of the species-specificCandida nest-ed PCRs were all 100%. Indeed, the inner primers designed fora given Candida species never amplified DNA from any of theother four Candida species (Fig. 1 and Table 2) or from any ofthe other fungal, bacterial, or parasitic species tested. No cross-reactivity with human DNA was observed (Table 2).

Sensitivity of C. albicans nested PCR applied to whole bloodand serum, compared to that of quantitative blood cultures inthe rabbit model. In the five rabbits injected with 108 CFU ofC. albicans ATCC 2091, results of quantitative blood culturesshowed that 90% of the microorganisms present in the blood1 min after injection were cleared within 4 min. Blood culturesthen remained positive, with a low concentration of C. albicans(1 to 17 CFU/ml) during the remaining 115 min of the exper-iment. Nested PCRs performed on whole blood were positivein 26 of 30 (86%) samples tested, comprising all of the 11 sam-ples in which the counts of C. albicans were greater than 10CFU/ml, and 15 of the 19 samples in which the counts wereequal to or lower than 10 CFU/ml (Fig. 2). When PCR was per-formed on serum, a total of 28 of 30 (93%) samples werepositive. The two negative serum samples were from the samerabbit and were drawn at 1 and 5 min after injection, when cellconcentrations of C. albicans were high (Fig. 2). NegativePCRs were not due to the presence of inhibitors in whole-blood or serum samples, since the internal positive PCR con-trols were always positive (Fig. 3).

When a smaller inoculum of 105 CFU of C. albicans ATCC2091 was injected, a candidemia of brief duration was observedin five rabbits and no candidemia was observed in one rabbit at1 min postinjection, and the PCRs performed on whole bloodand on sera were positive in 4 of 6 and 3 of 6 of these rabbits,respectively (Table 3). The blood cultures remained negative

FIG. 1. Sensitivities and specificities of DNA detection by species-specificnested PCRs of purified DNA from C. albicans, C. tropicalis, C. krusei, C. para-psilosis, and C. glabrata with ethidium bromide staining on agarose gel electro-phoresis. Shown are nested PCR products obtained with each species-specificinner pair of primers from different amounts of template DNA from the corre-sponding Candida species (lane 1, 1 pg; lane 2, 100 fg; lane 3, 10 fg; lane 4, 1 fg;lane 5, 0.1 fg). The size of the specific amplified fragment is indicated on the left.The specificity of the C. albicans primers was tested on 100 ng of purified DNAfrom C. tropicalis (lane 6), C. glabrata (lane 7), C. krusei (lane 8), and C. parap-silosis (lane 9). The specificity of the C. tropicalis primers was tested on 100 ng ofpurified DNA from C. albicans (lane 6), C. glabrata (lane 7), C. krusei (lane 8),and C. parapsilosis (lane 9). The specificity of the C. krusei primers was tested on100 ng of purified DNA from C. albicans (lane 6), C. tropicalis (lane 7), C. gla-brata (lane 8), and C. parapsilosis (lane 9). The specificity of the C. parapsilosisprimers was tested on 100 ng of purified DNA from C. albicans (lane 6), C. tropi-calis (lane 7), C. glabrata (lane 8), and C. krusei (lane 9). The specificity of theC. glabrata primers was tested on 100 ng of purified DNA of C. albicans (lane 6),C. tropicalis (lane 7), C. krusei (lane 8), and C. parapsilosis (lane 9). In theseexperiments, the amplified products generated by the Candida universal primers(ITS1 and ITS4) in the first reaction are sometimes visible. M, molecular weightmarker.

TABLE 2. Specificity of Candida species nested PCR

SpeciesNo. of positive PCRs/no. of DNA specimens

C. albicans C. tropicalis C. parapsilosis C. glabrata C. krusei

C. albicans 18/18a 0/18 0/18 0/18 0/18C. tropicalis 0/3 3/3a 0/3 0/3 0/3C. parapsilosis 0/3 0/3 3/3a 0/3 0/3C. glabrata 0/6 0/6 0/6 6/6a 0/6C. krusei 0/2 0/2 0/2 0/2 2/2a

A. fumigatus 0/5 0/5 0/5 0/5 0/5C. neoformans 0/1 0/1 0/1 0/1 0/1P. carinii 0/2 0/2 0/2 0/2 0/2M. canis 0/2 0/2 0/2 0/2 0/2T. rubrum 0/1 0/1 0/1 0/1 0/1T. gondii 0/1 0/1 0/1 0/1 0/1P. mirabilis 0/2 0/2 0/2 0/2 0/2E. coli 0/2 0/2 0/2 0/2 0/2S. aureus 0/2 0/2 0/2 0/2 0/2E. cloacae 0/1 0/1 0/1 0/1 0/1M. tuberculosis 0/1 0/1 0/1 0/1 0/1Human 0/5 0/5 0/5 0/5 0/5

a Homospecific interactions.

VOL. 37, 1999 SYSTEMIC CANDIDIASIS DIAGNOSIS BY PCR IN SERUM 927

on June 30, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

thereafter, but the PCRs performed on whole blood were stillpositive in two rabbits and one rabbit at 5 and 15 min post-injection, respectively (Table 3). In addition, the PCRs per-formed on sera were positive for 11 of the 35 samples drawnfrom 5 to 150 min postinjection but were always negative forsamples drawn 180 min after injection. Overall, among the 40culture-negative samples drawn later than 1 min after injec-tion, 3 (7%) were positive when PCR was performed on wholeblood and 11 (27%) were positive when PCR was performedon serum (P , 0.05%).

Sequence comparison. The sequences of the 386-bp frag-ments amplified from C. albicans ATCC 2091 and from thesera of infected rabbits were strictly homologous. They differedfrom the corresponding GenBank C. albicans sequence (acces-sion no. L47111) only by the insertion of a G base between po-sitions 353 and 354 (99.7% homology).

Purified Candida DNA clearance from rabbit blood usingnested PCR. When purified C. albicans DNA was directly in-jected into rabbits at a dose of 200 mg, the PCRs were positivein serum samples from 1 to at least 120 min. They were positivefrom 1 to at least 30 min when only 20 mg was injected (Table4).

DISCUSSION

We designed a nested Candida PCR assay in which an am-plified fragment of the Candida ITS repeated region was usedas a template for five different inner primer pairs, which werechosen for specific amplification of the DNAs of the five spe-cies most frequently causing human candidiasis (2, 28). The invitro specificity of our five PCRs was 100%. The sensitivity wassimilar to that published elsewhere for PCR targeting repeatedgenes and using specific probe hybridization assays for speciesdifferentiation (8, 12).

For in vivo evaluation, we used an experimental model ofinfection in rabbits in which candidemia was studied over aperiod of 120 min after injection of a 108-CFU C. albicansbolus. This inoculum was similar to that used in a previouslydescribed model of experimental candidemia in rabbits (1).Because of the large volumes of the blood samples that couldbe drawn from rabbits, we could precisely compare the sensi-tivities of our nested PCR in whole blood and serum. Such a

strict comparison had not and probably could not have beenperformed in studies of experimental candidiasis in smallerlaboratory animals (6, 15, 17). Compared to quantitative bloodcultures, the overall observed sensitivity of our PCR assay was86% for whole blood and 93% for serum. The PCR assay wasnegative in four whole-blood and two serum samples whichwere positive in culture. The sampling times at which the PCRwas negative in whole blood differed from those at which it wasnegative in serum, possibly due to the different origins of thetemplate DNA. Considering, first, that template DNA in wholeblood originated from DNA extracted ex vivo from Candidacells circulating in the blood and, second, that the four PCRswith false-negative results in whole blood were performed onsamples with low Candida counts (5, 4, 2, and 2 CFU/ml),negativity may reflect difficulty in extracting DNA ex vivo bycell lysis, as reported elsewhere (23, 24). PCR was positive in28 of 30 serum samples for which there was no such ex vivo cell

FIG. 3. Coamplification of internal PCR control and rabbit blood samples.Lanes 1, 2, 3, and 4, whole-blood samples; lanes 5 and 6, serum samples whichexhibited a negative C. albicans nested PCR; lanes 7 and 8, two whole-bloodsamples for which this PCR was positive. Amplification of the internal control isindicated by the presence of a 491-bp fragment, and amplification of C. albicansDNA in blood is indicated by the presence of a 386-bp fragment. M, molecularweight marker.

FIG. 2. Sensitivity of quantitative blood cultures compared to that of nested PCR performed on whole blood and on serum from five rabbits infected with 108 CFUof C. albicans. Each rabbit is represented by a circle at each sampling time. E, positive nested PCR in both whole blood and serum; K, positive nested PCR in wholeblood and negative nested PCR in serum; L, negative nested PCR in whole blood and positive nested PCR in serum.

928 BOUGNOUX ET AL. J. CLIN. MICROBIOL.

on June 30, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

lysis included in the protocol, suggesting that DNA could bephysiologically released in serum, which is in agreement withresults published by others (5, 6, 15). The only two negativeserum samples were from the same animal and were drawnearly, at 1 and 5 min after bolus injection, suggesting that theamount of DNA physiologically released during the first min-utes after injection was too small to be detectable by PCR.

In previously published clinical studies evaluating the sensi-tivity of PCR assays in the diagnosis of candidemia, samples forPCR were drawn at the same time as blood cultures and werefrozen, and only those yielding positive cultures were laterassayed by PCR (4, 10, 15). These samples were either wholeblood (4, 10) or serum (15), but the two kinds of samples werenever tested in the same study. In one study, the sensitivity ofa C. albicans- and C. glabrata-specific nested PCR performedon whole blood was 90% (19 of 21 positive blood cultures test-ed) (4). In another study, in which a C. albicans-specific probewas used for the detection of Candida DNA amplified fromwhole blood, all samples tested were positive by PCR, but theyall had Candida counts equal to or greater than 20 CFU/ml(10). In our work the only four false-negative PCR results onwhole blood were from samples with Candida cell counts below10 CFU/ml. By contrast, the 93% sensitivity that we found forPCR performed on serum samples from candidemic rabbitswas higher than the 79% sensitivity reported by others usinga single-copy gene (actin gene) target for a PCR assay per-formed on serum samples from candidemic patients (15). Weconcluded from this first set of experiments in rabbits injectedwith a large Candida inoculum (108 CFU) that the PCR assaythat we designed had a high sensitivity for detection of DNAboth in whole-blood samples and in serum samples drawnduring the culture-positive candidemic periods.

When we injected the rabbits with only 105 CFU of C. albi-cans, blood cultures were positive only during the 1st minfollowing injection, but PCRs performed on whole blood andon sera remained positive for a longer period, confirming thatPCR was more sensitive than cultures in diagnosing candi-demia, as reported both for murine (6, 15, 17, 24) and forhuman (5, 6, 15) candidemia. However, we also showed that

during the culture-negative period, the PCRs performed onsera remained positive longer than those performed on wholeblood. This could be explained by the different origins of tem-plate DNA amplified in the two types of samples. We suggestthat the positivity of whole-blood PCRs performed on culture-negative samples was due to the presence of noncultivableCandida cells, as previously reported (17, 23). The positivity ofPCR on serum samples long after cultures and PCR on wholeblood had become negative suggested that the clearance offree DNA was slower than that of either cultivable or noncul-tivable Candida cells. A slow clearance of free DNA was alsoobserved in the rabbits that we injected with purified C. albi-cans DNA.

In conclusion, our results showed that serum samples shouldbe used preferentially over whole blood to diagnose candi-demia by PCR. They also confirmed that Candida templateDNA which can be detected by PCR during the candidemicepisodes corresponds both to DNA from intact cultivable ornoncultivable Candida cells, and to free DNA released in vivo.Whether the same will be observed in neutropenic and post-surgery patients who are at high risk of Candida infection isnow being investigated in a prospective clinical trial.

ACKNOWLEDGMENTS

This work was supported in part by grant CRC95238 from Assis-tance Publique—Hopitaux de Paris and by a grant from the LigueNationale Franc(aise contre le Cancer.

REFERENCE

1. Baine, W. B., M. G. Koenig, and J. S. Goodman. 1974. Clearance of Candidaalbicans from the bloodstream of rabbits. Infect. Immun. 10:1420–1425.

2. Barns, S. M., D. J. Lane, M. L. Sogin, C. Bibeau, and W. G. Weinsburg. 1991.Evolutionary relationships among pathogenic Candida species and relatives.J. Bacteriol. 173:2250–2255.

2a.Bougnoux, M. E., C. Dupont, J. Mateo, P. Saulnier, D. Payen, and M. H.Nicolas-Chanoine. 1997. Rapid diagnosis of candidemia by DNA amplifica-tion applied to serum abstr. D-131, p. 106. In Abstracts of the 37th Inter-science Conference on Antimicrobial Agents and Chemotherapy. AmericanSociety for Microbiology, Washington, D.C.

3. Buchman, T. G., M. Rossier, W. G. Merz, and P. Charache. 1990. Detectionof surgical pathogens by in vitro DNA amplifications. Part 1. Rapid identi-fication of Candida albicans by in vitro amplification of a fungus-specificgene. Surgery 108:338–347.

4. Burgener-Kairuz, P., J. P. Zuber, P. Jaunin, T. G. Buchman, J. Billie, and M.Rossier. 1994. Rapid detection and identification of Candida albicans andTorulopsis (Candida) glabrata in clinical specimens by species-specific nestedPCR amplification of a cytochrome P-450 lanosterol-a-demethylase (L1A1)gene fragment. J. Clin. Microbiol. 32:1902–1907.

5. Burnie, J. P., N. Golbang, and R. C. Matthews. 1997. Semiquantitativepolymerase chain reaction enzyme immunoassay for diagnosis of dissemi-nated candidiasis. Eur. J. Clin. Microbiol. Infect. Dis. 16:346–350.

6. Chryssanthou, E., B. Andersson, B. Petrini, S. Lofdahl, and J. Tollemar.1994. Detection of Candida albicans DNA in serum by polymerase chainreaction. Scand. J. Infect. Dis. 26:479–485.

7. Crampin, A. C., and R. C. Matthews. 1993. Application of the polymerasechain reaction to the diagnosis of candidosis by amplification of an HSP 90gene fragment. J. Med. Microbiol. 39:233–238.

8. Einsele, H., H. Hebart, G. Roller, J. Loffler, I. Rothenhofer, C. A. Muller,R. A. Bowden, J.-A. van Burik, D. Engelhard, L. Kanz, and U. Schumacher.1997. Detection and identification of fungal pathogens in blood by usingmolecular probes. J. Clin. Microbiol. 35:1353–1360.

9. European Organization for Research and Treatment of Cancer and Inter-national Antimicrobial Therapy Cooperative Group. 1989. Empiric antifun-gal therapy in febrile granulocytopenic patients. Am. J. Med. 86:668–672.

10. Flahaut, M., D. Sanglard, M. Monod, J. Bille, and M. Rossier. 1998. Rapiddetection of Candida albicans in clinical samples by DNA amplification ofcommon regions from C. albicans-secreted aspartic proteinase genes. J. Clin.Microbiol. 36:395–401.

11. Fraser, V. J., M. Jones, J. Dunkel, S. Storfer, G. Medoff, and W. C. Dunagan.1992. Candidemia in a tertiary care hospital: epidemiology, risk factors, andpredictors of mortality. Clin. Infect. Dis. 15:414–421.

12. Fujita, S. I., B. A. Lasker, T. J. Lott, E. Reiss, and C. J. Morrison. 1995.Microtitration plate enzyme immunoassay to detect PCR-amplified DNAfrom Candida species in blood. J. Clin. Microbiol. 33:962–967.

13. Haynes, K. A., T. J. Westerneng, J. W. Fell, and W. Moens. 1995. Rapid

TABLE 3. C. albicans detection from quantitative blood culturesand from nested PCR performed on whole-blood and serum

samples of rabbits infected with 105 CFU of C. albicans

Method

No. positive/no. tested at the followingsampling time (min):

1 5 15 30 60 120 150 180 Total

Blood culture 5/6a 0/6 0/6 0/6 0/6 0/6 0/5 0/5 5/46PCR, whole blood 4/6 2/6 1/6 0/6 0/6 0/6 0/5 0/5 7/46PCR, serum 3/6 1/6 3/6 3/6 3/6 0/6 1/5 0/5 14/46

a Individual counts were 20, 8, 2, 2, 1, and 0 CFU/ml.

TABLE 4. PCR evaluation of serum clearance of different amountsof purified C. albicans DNA injected intravenously into four rabbits

Rabbitno.

Quantityof purifiedDNA (mg)

Resulta at the following samplingtime (min):

0 1 5 15 30 60 90 120 150 180

1 200 2 1 1 1 1 1 NA 1 NA NA2 200 2 1 1 1 1 1 1 1 NA 23 20 2 1 1 1 1 2 2 1 2 24 20 2 1 1 1 1 2 2 2 2 2

a NA; not available; 1, positive nested PCR; 2, negative nested PCR.

VOL. 37, 1999 SYSTEMIC CANDIDIASIS DIAGNOSIS BY PCR IN SERUM 929

on June 30, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from

detection and identification of pathogenic fungi by polymerase chain reac-tion amplification of large subunit ribosomal DNA. J. Med. Vet. Mycol. 33:319–325.

14. Jordan, J. A. 1994. PCR identification of four medically important Candidaspecies by using a single primer pair. J. Clin. Microbiol. 32:2962–2967.

15. Kan, V. L. 1993. Polymerase chain reaction for the diagnosis of candidemia.J. Infect. Dis. 168:779–783.

16. Karp, J. E., W. G. Merz, and P. Churache. 1991. Response to empiricamphotericin B during antileukemic therapy-induced granulocytopenia.Rev. Infect. Dis. 13:592–599.

17. Makimura, K., S. Y. Murayama, and H. Yamaguchi. 1994. Detection of awide range of medically important fungi by the polymerase chain reaction.J. Med. Microbiol. 40:358–364.

18. Pfaller, M. A. 1996. Nosocomial candidiasis: emerging species, reservoirs,and modes of transmission. Clin. Infect. Dis. 22:S89–S94.

19. Pittet, D., M. Monod, P. M. Suter, E. Frenk, and R. Auckenthaler. 1994.Candida colonization and subsequent infections in critically ill surgical pa-tients. Ann. Surg. 220:751–758.

20. Pizzo, P. A. 1993. Management of fever in patients with cancer and treat-ment-induced neutropenia. N. Engl. J. Med. 328:1323–1332.

21. Robert, F., F. Lebreton, M.-E. Bougnoux, A. Paugam, D. Wassermann, M.Schlotterer, C. Tourte-Schaefer, and J. Dupouy-Camet. 1995. Use ofrandom amplified polymorphic DNA as a typing method for Candidaalbicans in epidemiological surveillance of a burn unit. J. Clin. Microbiol.33:2366–2371.

22. Shin, J. H., F. S. Notle, and C. J. Morrison. 1997. Rapid identification of

Candida species in blood cultures by a clinically useful PCR method. J. Clin.Microbiol. 35:1454–1459.

23. van Deventer, A. J. M., W. H. F. Goessens, A. van Belkum, H. J. A. van Vliet,E. W. M. van Etten, and H. A. Verbrugh. 1995. Improved detection ofCandida albicans by PCR in blood of neutropenic mice with systemic can-didiasis. J. Clin. Microbiol. 33:625–628.

24. van Deventer, A. J. M., W. H. F. Goessens, A. van Belkum, J. A. van Vliet,W. M. van Etten, and H. A. Verbrugh. 1996. PCR monitoring of response toliposomal amphotericin B treatment of systemic candidiasis in neutropenicmice. J. Clin. Microbiol. 34:25–28.

25. Weinstein, M. P., M. L. Towns, S. M. Quartey, S. Mirrett, L. G. Reimer, G.Parmigiani, and L. Barth Reller. 1997. The clinical significance of positiveblood cultures in the 1990s: a prospective comprehensive evaluation of themicrobiology, epidemiology, and outcome of bacteremia and fungemia inadults. Clin. Infect. Dis. 24:584–602.

26. Wey, S. B., M. Mori, M. A. Pfaller, R. F. Woolson, and R. P. Wenzel. 1988.Hospital-acquired candidemia: the attributable mortality and excess lengthof stay. Arch. Intern. Med. 148:2642–2645.

27. White, T. J., T. D. Burns, S. B. Lee, and J. W. Taylor. 1990. Amplificationand direct sequencing of fungal ribosomal RNA genes for phylogenetics, p.315–322. In M. A. Innis, D. H. Gelfand, J. J. Sninsky, and T. J. White (ed.),PCR protocols. A guide to methods and applications. Academic Press, Inc.,San Diego, Calif.

28. Wingard, J. R. 1995. Importance of Candida species other than C. albicansas pathogens in oncology patients. Clin. Infect. Dis. 20:115–125.

29. Wingard, J. R. 1994. Infections due to resistant Candida species in patientswith cancer who are receiving chemotherapy. Clin. Infect. Dis. 19:S49–S53.

930 BOUGNOUX ET AL. J. CLIN. MICROBIOL.

on June 30, 2020 by guesthttp://jcm

.asm.org/

Dow

nloaded from