Article
Vol. 11. No. 8
693
Eur. J. Clin Mierobiol. Infect. Dis., August 1992, p. 693-697 0934-9723/92/08 0693-05 $ 3.00/0
Efficacy of Single-Dose Azithromycin versus Doxycycline in the Treatment of Cervical Infections Caused by Chlamydia
trachomatis
J.M. O s s e w a a r d e 1 . , E H . E P l a n t e m a 2, M. R i e f f e 1, R.P. N a w r o c k i 1, A . d e V r i e s 1, A.M. van Loon I
The efficacy of single-dose azithromycin therapy in the treatment of cervical Chlamydla trachomatls infections was compared to that of a standard seven-day course of treatment with doxycycline. Cervical samples from 60 patients reacted positively in an enzyme immunoassay for detection of Chlamydia trachomatis. In 31 patients Chlamydia trachomatts was isolated from the sample taken before treatment. Fourteen of the 31 patients were treated with doxycycline and 17 with azithromycin. All cultures of samples taken one and four weeks after the start of therapy were negative. All 31 isolates showed a similar pattern of MICs for the seven antibiotics tested, including azithromycin and doxycycline. No differences were observed between isolates of different serovars. In samples from four patients chlamydial D N A could be detected by PCR one week after the start of the therapy and in two patients also after four weeks. No difference in microbiological parameters could be observed between the two treatment groups. It is concluded that single-dose azithromycin is as effective as a seven-day course of doxycycline in the therapy of cervical Chlamydia trachomatis infections.
The standard therapy for Chlamydia trachomatis infections consists of a 7 to 14 day course of a tetracycline, usually doxycycline. Experience in the treatment of gonorrhea patients has shown that due to a frequent lack of patient compliance single-dose therapy is much more reliable than a course of several days. Until the introduction of azithromycin no appropriate agent for singledose therapy of Chlamydia trachomatis infections was available. Azithromycin, a new macrolide antibiotic, has pharmacological properties, in particular high prolonged tissue concentrations, which suggest it would be suitable for single-dose therapy (1). In this study we compared a singledose of azithromycin with a seven-day course of doxycycline in the therapy of patients with cervical Chlamydia trachomatis infections. The presence and viability of Chlamydia trachomatis was d e t e r m i n e d by the polymerase chain reaction
1Laboratory of Virology, National Institute of Public Health and Environmental Protection, PO Box 1, 3720 BA Bilthoven, The Netherlands. 2Carolus-Liduina Hospital, PO Box 10100, 5280 GA Boxtel, The Netherlands.
(PCR) and cell culture. In addition, we determined the MICs o f azithromycin and six other antibiotics for 31 clinical isolates.
Patients and Methods
Patients and Samples. In conjunction with a multi-center trial to determine the clinical efficacy of single-dose azithromycin therapy, female patients with symptoms of a sexually transmitted disease were asked to participate in this study. An enzyme immunoassay was used to detect Chlamydia trachomatis in cervical samples. From each patient in whom Chlamydia trachomatis was detected, three samples were taken for culture and PCR, the first before the start of antibiotic treatment, the second one week later and the third four weeks after the start of the treatment. All cervical swabs were placed in 4SP chlamydia transport medium (0.4 M sucrose-phosphate buffer of pH 7.2 supplemented with 10 % fetal calf serum and antibiotics) and stored at -70 *C until cultured. All patients were asked to refrain from sexual intercourse until the second sample had been taken. Patients were assigned in a predetermined random order to one of two treatment schedules: azithromycin in a single dose of 1 g or doxycycline in a dosage of 200 mg daily for seven days.
694
Culture of Chlamydia trachornatis. HeLa 229 cells were used for isolation of Chlamydia trachomatis from clinical samples and for propagation of isolated strains. The HeLa 229 cells were maintained in Iseove's Modified Dulbeeeo's medium (Gibco, USA) supplemented with 10 % fetal calf serum and antibiotics. Clinical samples and propagated strains were inoculated onto one-day-old DEAE-dextran treated monolayers and centrifuged at 4800 x g for 1 h at 25 *C. The cultures were incubated at 37 *C in 5 % CO2 for three days. All samples were inoculated onto one monolayer in a 24-well microtiter plate (Nune, Denmark) and growth was assessed by staining with fluorescein labelled monoclonal antibodies (Microq~rak, Syva, USA). All samples were also inoculated onto monolayers on glass cover slips in one or more flat bottom tubes (Greiner, Germany) for serial passaging and propagation. All clinical samples were passaged once blindly, Monotayers with more than 75 % infected cells were sonicated and stored in 4SP medium at -70 *C. Determination of Minimal Inhibitory Concentrations (MICs). All clinical isolates were passaged two or three times in Iscove's medium free of antibiotics. One hundred ~ti of an appropriate dilution in Iscove's medium to yield between 5 and 50 inclusions per microscopic field at 192 (6 x 32) times magnification was inoculated onto a oneday-old DEAE-dextran treated monolayer of antibiotic free HeLa 229 cells in each welt of a 96-well microtiter plate (Nunc). After eentrifugation for 1 h at 4800 x g, 100 ~1 of the antibiotic dilution was added to the inoculum in the wells and the plates were incubated at 37 *C in 5 % CO2 for three days. The following range of final concentrations of antibiotics was used in a twofold dilution series: polymyxin 0.08-160 lag/ml; ofloxacin 0.01-16~tglml; roxithromyein, doxycycline, azithromycin, tetracycline and minocyeline 0.001-2/~g/ml. The dilutions of the antibiotics were prepared as follows. Polymyxin was dissolved in 0.01 M phosphate buffered saline (PBS) of pH 7.2 at 3.2 mgtml and diluted ten times with Iscove's medium. Doxycycline and tetracycline were dissolved in PBS at 0.32 mg/ml and diluted 80 times with Iscove's medium. Minocycline was dissolved in PBS at 0.96 mg/ml and diluted 240 times with Iscove's medium. Roxithromycin and azithromycin were dissolved in ethanol at 6.4 mg/ml, diluted twice with PBS and then 800 times with Iscove's medium. Ofloxacin was dissolved in acetic acid at 32 mg/ml, diluted 10 times with PBS and then 100 times with Iscove's medium. Two wells without antibiotics were used as control. One dilution series of each antibiotic without inoculum was included to assess cell toxicity. After three days the monolayers were stained with fluorescein labelled monoclonal antibodies (Syva). The lowest concentration of the antibiotic, which completely inhibited the formation of inclusions, was considered to be the MIC.
Serovar Typing. Serovar typing was performed with a set of monoelonai antibodies developed at our own laboratory and eapable of distinguishing all 15 serovars of Chlamydia trachomatis. The typing was done in a dot enzyme immunoassay with antigen derived from flat bottom tubes with more than 75 % infected cells (2).
Oligonucleotides. The sequences of the primers for amplification (primer I:5'-CTAGGCG'VI~GTACTCCGTCA-3' and primer 2:5'-TCCTCAGAAG'VITA-
E u r . J. Clin. M i c r o b i o l , Infect. Dis.
TGCACT-3') were kindly provided by Organon Teknika, The Netherlands, who amplified a fragment of 200 base pairs of the ORF 2 of the chlamydial plasmid (3)..A database search revealed no other sequences with significant homology. An internal oligonu¢leotide (5'C G G T F G C T C G A A G C A C G T G C G - 3 ' ) was chosen as probe for spot hybridization. The probe was labelled with digoxigenin-dUTP using terminal transferase according to the instructions of the manufacturer (Boehringer, Germany).
Polymerase Chain Reaction (PCR). All available samples of the 31 culture positive patients were tested by PCR for the presence of ehlamydial DNA. From each sample 0.6 ml was centrifuged for 30 min at 12,500 x g in a mierofuge. The pellet was suspended in 25/al lysis buffer (50 mM TRIS-HCI, pH 7.5; 1 % Triton X-100; 1 mM EDTA; 400/ag/ml proteinase K) using a pellet pestle (Kontes, USA). After incubation at 37 *C for 1 h the lysates were boiled for 10 min and briefly centrifuged. From each lysate 2 ~1 was added to 28 ~tl of the PCR mixture (50 mM TRIS-HCI, pH 8.5 at 37"C; 50 mM NaCI; 6 mM MgCIz; 2 mM Dq'T; 1.3 units of AmpliTaq [Cetus, USA]; 1 mM of each NT; 1 laM of each primer). The samples were cycled 30 times in a Thermal Cycler (Perkin Elmer, USA) (one cycle of I min 30 sec at 94 *C; 1 see. at 55 *C; 2 min 30 see at 72 *C; followed by 29 cycles of 30 see at 94 *C; 1 see at 55 *C; 2 rain 30 see at 72 *C). Chlamydial DNA was detected by electrophoresis of I0 lal of the amplicon in 2 % agarose (SeaKem GTG agarose, FMC, USA). A clearly visible band of 200 base pairs in the gel after staining with ethidium bromide was considered a positive reaction. DNA was also detected by a spot hybridization technique as follows. After alkaline denaturation and neutralization with ammonium acetate 10 /al of the amplieon was spotted onto Zeta-Probe (BioRad, USA) using a HybriDot Manifold (BRL, USA). The filters were washed, air dried and baked in a microwave oven. The filters were first incubated with prehybridization buffer (25 mM sodium phosphate buffer pH 6.5; 5x SSC; 50 % formamide; 5x Denhardt; 0.1% SDS; 400 ~tg/ml denatured salmon testes DNA) for 2--4 h at 42 *C. The filters were then hybridized overnight at 42 *C with 50 ng/ml digoxigenin labelled internal oligonueleotide in hybridization buffer (25 mM sodium phosphate buffer pH 6.5; 5x SSC; 45 % formamide; lxDenhardt; 0.1% SDS; 10 % dextransulphate; 100 lag/ml denatured salmon testes DNA). The filters were washed four times for 2 min at 42 *C with 2x SSC with 0.1% SDS, two times for 2 min at 42 *C with 0.2x SSC with 0,1% SDS, two times for 15 rain at 65 *C with 0.16x SSC with 0.1% SDS and once for 1 min at 42 *C with 2x SSC with 0.1% SDS. The filters were blocked, incubated with anti-digoxigenin antibodies labelled with alkaline phosphatase and developed according to the instructions of the manufacturer (Boehringer, Germany). A clearly visible spot was considered a positive reaction.
Statistical Analysis. The MIC results were analyzed by univariate and multivariate analysis of variance using the Complete Statistical System PC software package from StatSoft. A result was considered significant when the corresponding p value was less than 0.05. Confidence intervals of fraction estimates were calculated using approximation with a normal distribution.
Vol. 11,1992
695
Results
All patients indicated complete compliance with the therapy and refrained from sexual intercourse until the second sample was taken for culture.
Chlamydia trachomatis was isolated from 31 of 60 samples taken b e f o r e treatment. All isolates could be p r o p a g a t e d in three to five passages to more than 75 % infected cells. All samples taken one week and four weeks after treatment were negative on culture. T h e results of serovar typing of the 31 clinical isolates were as follows: 12 isolates belonged to serovar D; 10 to serovar E, 7 to serovar F and 2 to serovar J. Table 1 shows the results of determination of the MICs. Polymyxin and ofloxacin were significantly less active than the o t h e r five antibiotics. The difference b e t w e e n the MICs of azithromycin and tetracycline was not statistically significant, but the M I C of roxithromycin, minocycline and doxycycline was significantly lower than that of
azithromycin and tetracycline. T h e r e were no differences in the MICs attributable to the serovar. In addition, there was no significant difference between the MICs of the clinical isolates of the two treatment groups. After amplification by the PCR, chlamydial D N A was detected by both gel electrophoresis and spot hybridization in all 30 available samples t a k e n before the start of the treatment. Two of the 30 samples taken one week after the start o f the therapy were D N A positive o n gel electrophoresis. Both patients received azithromycin. Spot hybridization detected two additional D N A positive samples: one from a patient in the azithromycin group and one from a patient in the doxycycline group. O n e sample from a patient in the doxycycline group was not available for D N A detection. T h e first and third samples from this patient were both D N A positive. T h e o t h e r patient with a D N A positive sample taken four weeks after the start of the therapy received azithromycin. Table 2 shows the details of the D N A detection results. T h e serovar distribution
Table 1: Minima~inhibit~rycon~entrati~nsf~ra~3~is~ates~fCh~amydiatrach~matisandf~rthef~urser~vars. Antibiotic
MIC (gg/ml) Range (n = 31)
Polymyxin Ofloxaein Tetracycline Azithromycin Doxycycline Minocycline Roxithrornycin
40- > 160 0.5 - 1.0 0.03- 0.06 0.01 - 0.06 0.01- 0.02 0.01-0.03 0.01 - 0.03
Mean MIC (pg/ml) for serovar
Mean (n = 31)
Type D (n = 12)
Type E (n = t0)
Type F (n = 7)
Type J (n = 2)
NC* 0.516 0.034 0.029 0.019 0.017 0.017
NC* 0.500 0.030 0.030 0.019 0.018 0.015
NC* 0.500 0.033 0.030 0.020 0.019 0.018
NC* 0.571 0.030 0.031 0.019 0.013 0.016
NC* 0.500 0.060 0.015 0.020 0.020 0.020
*NC = Not calculated as MICs exceeded 160 ~ag/mt. Table 2: Numbers of positive results in culture and PCR for detection of Chlamydiatrachomatis.
Sample collection
Culture
Before treatment
PCR Electrophoresis
Spot hybridization
31/31 (100 %)
30/30~ (100 %)
30/30a (100 %)
One week after treatment
0/31
(0 %)
2/30b
(7 %)
4/30b (13 %)
Four weeks after treatment
0/29¢
(0 %)
2/29¢
(7 %)
2/29c
(7 %)
aThe first sample of one patient was not available for PCR; the second and third sample of this bPatient were negative by PCR. The second sample of one patient was not available for PCR; the first and third sample of this patient were positive by gel electrophoresis and spot hybridization. c No third samples were available from two patients for culture or PCR.
696
of the isolates of the patients with a D N A positive sample one or four weeks after therapy were: two isolates of serovar D, two isolates of serovar E and one isolate of serovar J.
Discussion
Samples were obtained from 60 patients in whom a previous sample had reacted positively in an enzyme immunoassay for detection of Chlamydia trachomatis antigen at a local laboratory. However, Chlamydia trachomatis was isolated from only 31 of these patients. During the study period the local laboratory performing the assays had experienced some technical difficulties resulting in a number of false positive results. This explains the apparently low recovery rate of our cultures. The problem of reduced performance of chlamydia detection assays in non-research laboratories has also been experienced in other countries (4). Using these 31 isolates the MIC values of seven antibiotics were determined. The method used is a modification of the method outlined by Ehret and Judson (5). Since the growth of many strains is limited in Eagle's minimal essential medium (6) we used Iscove's modified Dulbecco's medium. We did not include cycloheximide in the medium since it interferes with the production of normal infectious elementary bodies (7) and might therefore interfere with the action of the antibiotics. Since we were interested in obtaining the MIC values only we used microtiter plates for convenience. They were centrifuged at the same speed as the fiat bottom tubes and usually yielded the same number of inclusions per microscopic field as cover slips in flat bottom tubes. Polymyxin, included as a negative control, showed almost no activity against any of the clinical isolates. Ofloxacin, which so far has been shown to be the most active member of the fluoroquinolones (8), had an MIC of 0.5/ag/ml. Some studies indeed suggest this drug would have clinical efficacy (9). The other five drugs had MIC values that were very similar to each other. The MICs were within the range found previously (5, 10). The negative cultures after therapy showed that one week after therapy with a single dose of azithromycin or a seven-day course of doxycycline no viable Chlamydia trachomatis organisms were present at the sites of infection. Therefore both therapies eradicated live
Eur. J. Clin. Microbiol. Infect. Dis.
Chlamydia trachomatis organisms from these sites. However, chlamydial D N A could be detected by PCR in about 15 % (95 % confidence interval: 3 %-29 %) of the patients one week after the start of the therapy and in about 6 % (95 % confidence interval: 0 %-16 %) of the patients four weeks after therapy. The observation that four weeks after the start of therapy less samples were positive than after one week indicates that the presence of chlamydial D N A was a remnant of the infection rather than re-exposure. This is consistent with the fact that all patients stated they refrained from sexual intercourse until the second sample was taken for culture and with the negative cultures. In contrast with our finding Claas et al. (11) found no residual chlamydial D N A by the P C R one week after doxycycline therapy in 32 patients. However, using an antigen detection enzyme immunoassay Radcliffe et al. (12) found chlamydial antigen in 3 of 91 samples (95 % confidence interval: 0 % 7 %) one week after therapy. They concluded that this kind of method should not be used as a test of cure one week after therapy. Since the P C R is potentially much more sensitive than any antigen detection assay currently available and cannot discriminate between viable and non-viable microorganisms, we conclude that D N A detection by P C R is not suitable as a test of cure and cannot be used to monitor chlamydial infection. Further long term follow-up studies are needed to assess the value of the P C R in discriminating between D N A detected after a cured infection, in chronic infections and in acute (re)infections. Thus culture is still the only way to detect the presence of viable microorganisms and therefore the only way to study treatment failure or resistant strains (13). To detect possible differences in the MICs of serovars, all isolates were typed using monoclonal antibodies. The distribution of the serovars was similar to that found previously in the Netherlands (14), indicating that our study was not biased by an abnormal serovar distribution caused by selection of fast growing isolates or a local outbreak of one serovar. No difference was observed between MIC values of the different serovars. We found no differences between the two treatment groups with respect to the culture results, the MIC values, D N A detection and serovar types, and conclude that a single dose of azithromycin and a seven-day course of doxycycline are equally efficacious.
Vo1.11,1992
References 1. Foulds G, Shepard RM, Johnson RB: The pharmacokinetics of azithromyein in human serum and tissues. Journal of Antimicrobial Chemotherapy 1990, 25, Supplement A: 73-82. 2. Barnes RC, Wang SP, Kuo CC, Stamm WE: Rapid immunotyping of Chlamydia trachornatis with monoelonal antibodies in a solid phase enzyme immunoassay. Journal of Clinical Microbiology 1985, 22: 609613. 3. Sriprakash KS, Maeavoy ES: Characterization and sequence of a plasmid from the trachoma biovar of Chlamydia trachomatis. Plasmid 1987, 18: 205-214. 4. Taylor-Robinson D, Thomas BJ: Laboratory techniques for the diagnosis of chlamydial infections. Genitourinary Medicine 1991, 67: 256--266. 5. Ehret JM, Judson FN: Susceptibility testing of Chlamydia trachornatis: from eggs to monoclonal antibodies. Antimicrobial Agents and Chemotherapy 1988, 32: 1295-1299. 6. Johnson FWA, Hobson D: Factors affecting the sensitivity of replicating McCoy cells in the isolation and growth of Chlamydia A (TRIC agents). Journal of Hygiene 1976, 76: 441-451. 7. Karayiannis P, Hobson D, Lee N: Effect of cycloheximide on the infective yield of a genital strain of Chlarnydia trachomatis in McCoy cells. Infection and Immunity 1981, 33: 309-311.
697
8. Endtz HPH, Ossewaarde JM, van der Korput PJGM, Weiland HI':. In vitro activity of eight quinolones against Chlamydia trachomatis. Reviews of Infectious Diseases 1989, 11, Supplement 5: 1278. 9. Ronald AR, Peeling RW: Chlamydial infections and the quinolones. European Journal of Clinical Microbiology and Infectious Diseases 1991, 10: 351-354. 10. Seieux C, Bianchi A, Chappey B, Vassias I, Perol Y: In vitro activity of azithromycin against Chtamydia trachomatis. Journal of Antimierobial Chemotherapy 1990, 25, Supplement A: 7-10. 11. Claas HCJ, Wagenvoort JHT, Niesters HGM, "13oTT, van Rijsoort-Vos JH, Quint WGV: Diagnostic value of the polymerase chain reaction for ehlamydia detection as determined in a follow-up study. Journal of Clinical Microbiology 1991, 29: 42-45. 12. Radcliffe KW, Rowen D, Mereey DE, Mumlaz G, Ridgway GL, Robinson A J, Bingham JS: Is a test of cure necessary following treatment for cervical infection with Chlamydia trachomatis. Genitourinary Medicine 1990, 66:44 A,~..46. 13. Jones RB, van der Pol B, Marlin DH, Shepard MK: Partial characterization of Chlamydia trachomatis isolates resistant to multiple antibiotics, Journal of Infectious Diseases 1990, 162: 1309-1315. 14. Wagenvoort JIIT, Suchland RJ, Stamm WE: Serovar distribution of urogenital Chlamydia trachomatis strains in the Netherlands. Genitourinary Medicine 1988, 64: 159-161.
Vol. 11. No. 8
693
Eur. J. Clin Mierobiol. Infect. Dis., August 1992, p. 693-697 0934-9723/92/08 0693-05 $ 3.00/0
Efficacy of Single-Dose Azithromycin versus Doxycycline in the Treatment of Cervical Infections Caused by Chlamydia
trachomatis
J.M. O s s e w a a r d e 1 . , E H . E P l a n t e m a 2, M. R i e f f e 1, R.P. N a w r o c k i 1, A . d e V r i e s 1, A.M. van Loon I
The efficacy of single-dose azithromycin therapy in the treatment of cervical Chlamydla trachomatls infections was compared to that of a standard seven-day course of treatment with doxycycline. Cervical samples from 60 patients reacted positively in an enzyme immunoassay for detection of Chlamydia trachomatis. In 31 patients Chlamydia trachomatts was isolated from the sample taken before treatment. Fourteen of the 31 patients were treated with doxycycline and 17 with azithromycin. All cultures of samples taken one and four weeks after the start of therapy were negative. All 31 isolates showed a similar pattern of MICs for the seven antibiotics tested, including azithromycin and doxycycline. No differences were observed between isolates of different serovars. In samples from four patients chlamydial D N A could be detected by PCR one week after the start of the therapy and in two patients also after four weeks. No difference in microbiological parameters could be observed between the two treatment groups. It is concluded that single-dose azithromycin is as effective as a seven-day course of doxycycline in the therapy of cervical Chlamydia trachomatis infections.
The standard therapy for Chlamydia trachomatis infections consists of a 7 to 14 day course of a tetracycline, usually doxycycline. Experience in the treatment of gonorrhea patients has shown that due to a frequent lack of patient compliance single-dose therapy is much more reliable than a course of several days. Until the introduction of azithromycin no appropriate agent for singledose therapy of Chlamydia trachomatis infections was available. Azithromycin, a new macrolide antibiotic, has pharmacological properties, in particular high prolonged tissue concentrations, which suggest it would be suitable for single-dose therapy (1). In this study we compared a singledose of azithromycin with a seven-day course of doxycycline in the therapy of patients with cervical Chlamydia trachomatis infections. The presence and viability of Chlamydia trachomatis was d e t e r m i n e d by the polymerase chain reaction
1Laboratory of Virology, National Institute of Public Health and Environmental Protection, PO Box 1, 3720 BA Bilthoven, The Netherlands. 2Carolus-Liduina Hospital, PO Box 10100, 5280 GA Boxtel, The Netherlands.
(PCR) and cell culture. In addition, we determined the MICs o f azithromycin and six other antibiotics for 31 clinical isolates.
Patients and Methods
Patients and Samples. In conjunction with a multi-center trial to determine the clinical efficacy of single-dose azithromycin therapy, female patients with symptoms of a sexually transmitted disease were asked to participate in this study. An enzyme immunoassay was used to detect Chlamydia trachomatis in cervical samples. From each patient in whom Chlamydia trachomatis was detected, three samples were taken for culture and PCR, the first before the start of antibiotic treatment, the second one week later and the third four weeks after the start of the treatment. All cervical swabs were placed in 4SP chlamydia transport medium (0.4 M sucrose-phosphate buffer of pH 7.2 supplemented with 10 % fetal calf serum and antibiotics) and stored at -70 *C until cultured. All patients were asked to refrain from sexual intercourse until the second sample had been taken. Patients were assigned in a predetermined random order to one of two treatment schedules: azithromycin in a single dose of 1 g or doxycycline in a dosage of 200 mg daily for seven days.
694
Culture of Chlamydia trachornatis. HeLa 229 cells were used for isolation of Chlamydia trachomatis from clinical samples and for propagation of isolated strains. The HeLa 229 cells were maintained in Iseove's Modified Dulbeeeo's medium (Gibco, USA) supplemented with 10 % fetal calf serum and antibiotics. Clinical samples and propagated strains were inoculated onto one-day-old DEAE-dextran treated monolayers and centrifuged at 4800 x g for 1 h at 25 *C. The cultures were incubated at 37 *C in 5 % CO2 for three days. All samples were inoculated onto one monolayer in a 24-well microtiter plate (Nune, Denmark) and growth was assessed by staining with fluorescein labelled monoclonal antibodies (Microq~rak, Syva, USA). All samples were also inoculated onto monolayers on glass cover slips in one or more flat bottom tubes (Greiner, Germany) for serial passaging and propagation. All clinical samples were passaged once blindly, Monotayers with more than 75 % infected cells were sonicated and stored in 4SP medium at -70 *C. Determination of Minimal Inhibitory Concentrations (MICs). All clinical isolates were passaged two or three times in Iscove's medium free of antibiotics. One hundred ~ti of an appropriate dilution in Iscove's medium to yield between 5 and 50 inclusions per microscopic field at 192 (6 x 32) times magnification was inoculated onto a oneday-old DEAE-dextran treated monolayer of antibiotic free HeLa 229 cells in each welt of a 96-well microtiter plate (Nunc). After eentrifugation for 1 h at 4800 x g, 100 ~1 of the antibiotic dilution was added to the inoculum in the wells and the plates were incubated at 37 *C in 5 % CO2 for three days. The following range of final concentrations of antibiotics was used in a twofold dilution series: polymyxin 0.08-160 lag/ml; ofloxacin 0.01-16~tglml; roxithromyein, doxycycline, azithromycin, tetracycline and minocyeline 0.001-2/~g/ml. The dilutions of the antibiotics were prepared as follows. Polymyxin was dissolved in 0.01 M phosphate buffered saline (PBS) of pH 7.2 at 3.2 mgtml and diluted ten times with Iscove's medium. Doxycycline and tetracycline were dissolved in PBS at 0.32 mg/ml and diluted 80 times with Iscove's medium. Minocycline was dissolved in PBS at 0.96 mg/ml and diluted 240 times with Iscove's medium. Roxithromycin and azithromycin were dissolved in ethanol at 6.4 mg/ml, diluted twice with PBS and then 800 times with Iscove's medium. Ofloxacin was dissolved in acetic acid at 32 mg/ml, diluted 10 times with PBS and then 100 times with Iscove's medium. Two wells without antibiotics were used as control. One dilution series of each antibiotic without inoculum was included to assess cell toxicity. After three days the monolayers were stained with fluorescein labelled monoclonal antibodies (Syva). The lowest concentration of the antibiotic, which completely inhibited the formation of inclusions, was considered to be the MIC.
Serovar Typing. Serovar typing was performed with a set of monoelonai antibodies developed at our own laboratory and eapable of distinguishing all 15 serovars of Chlamydia trachomatis. The typing was done in a dot enzyme immunoassay with antigen derived from flat bottom tubes with more than 75 % infected cells (2).
Oligonucleotides. The sequences of the primers for amplification (primer I:5'-CTAGGCG'VI~GTACTCCGTCA-3' and primer 2:5'-TCCTCAGAAG'VITA-
E u r . J. Clin. M i c r o b i o l , Infect. Dis.
TGCACT-3') were kindly provided by Organon Teknika, The Netherlands, who amplified a fragment of 200 base pairs of the ORF 2 of the chlamydial plasmid (3)..A database search revealed no other sequences with significant homology. An internal oligonu¢leotide (5'C G G T F G C T C G A A G C A C G T G C G - 3 ' ) was chosen as probe for spot hybridization. The probe was labelled with digoxigenin-dUTP using terminal transferase according to the instructions of the manufacturer (Boehringer, Germany).
Polymerase Chain Reaction (PCR). All available samples of the 31 culture positive patients were tested by PCR for the presence of ehlamydial DNA. From each sample 0.6 ml was centrifuged for 30 min at 12,500 x g in a mierofuge. The pellet was suspended in 25/al lysis buffer (50 mM TRIS-HCI, pH 7.5; 1 % Triton X-100; 1 mM EDTA; 400/ag/ml proteinase K) using a pellet pestle (Kontes, USA). After incubation at 37 *C for 1 h the lysates were boiled for 10 min and briefly centrifuged. From each lysate 2 ~1 was added to 28 ~tl of the PCR mixture (50 mM TRIS-HCI, pH 8.5 at 37"C; 50 mM NaCI; 6 mM MgCIz; 2 mM Dq'T; 1.3 units of AmpliTaq [Cetus, USA]; 1 mM of each NT; 1 laM of each primer). The samples were cycled 30 times in a Thermal Cycler (Perkin Elmer, USA) (one cycle of I min 30 sec at 94 *C; 1 see. at 55 *C; 2 min 30 see at 72 *C; followed by 29 cycles of 30 see at 94 *C; 1 see at 55 *C; 2 rain 30 see at 72 *C). Chlamydial DNA was detected by electrophoresis of I0 lal of the amplicon in 2 % agarose (SeaKem GTG agarose, FMC, USA). A clearly visible band of 200 base pairs in the gel after staining with ethidium bromide was considered a positive reaction. DNA was also detected by a spot hybridization technique as follows. After alkaline denaturation and neutralization with ammonium acetate 10 /al of the amplieon was spotted onto Zeta-Probe (BioRad, USA) using a HybriDot Manifold (BRL, USA). The filters were washed, air dried and baked in a microwave oven. The filters were first incubated with prehybridization buffer (25 mM sodium phosphate buffer pH 6.5; 5x SSC; 50 % formamide; 5x Denhardt; 0.1% SDS; 400 ~tg/ml denatured salmon testes DNA) for 2--4 h at 42 *C. The filters were then hybridized overnight at 42 *C with 50 ng/ml digoxigenin labelled internal oligonueleotide in hybridization buffer (25 mM sodium phosphate buffer pH 6.5; 5x SSC; 45 % formamide; lxDenhardt; 0.1% SDS; 10 % dextransulphate; 100 lag/ml denatured salmon testes DNA). The filters were washed four times for 2 min at 42 *C with 2x SSC with 0.1% SDS, two times for 2 min at 42 *C with 0.2x SSC with 0,1% SDS, two times for 15 rain at 65 *C with 0.16x SSC with 0.1% SDS and once for 1 min at 42 *C with 2x SSC with 0.1% SDS. The filters were blocked, incubated with anti-digoxigenin antibodies labelled with alkaline phosphatase and developed according to the instructions of the manufacturer (Boehringer, Germany). A clearly visible spot was considered a positive reaction.
Statistical Analysis. The MIC results were analyzed by univariate and multivariate analysis of variance using the Complete Statistical System PC software package from StatSoft. A result was considered significant when the corresponding p value was less than 0.05. Confidence intervals of fraction estimates were calculated using approximation with a normal distribution.
Vol. 11,1992
695
Results
All patients indicated complete compliance with the therapy and refrained from sexual intercourse until the second sample was taken for culture.
Chlamydia trachomatis was isolated from 31 of 60 samples taken b e f o r e treatment. All isolates could be p r o p a g a t e d in three to five passages to more than 75 % infected cells. All samples taken one week and four weeks after treatment were negative on culture. T h e results of serovar typing of the 31 clinical isolates were as follows: 12 isolates belonged to serovar D; 10 to serovar E, 7 to serovar F and 2 to serovar J. Table 1 shows the results of determination of the MICs. Polymyxin and ofloxacin were significantly less active than the o t h e r five antibiotics. The difference b e t w e e n the MICs of azithromycin and tetracycline was not statistically significant, but the M I C of roxithromycin, minocycline and doxycycline was significantly lower than that of
azithromycin and tetracycline. T h e r e were no differences in the MICs attributable to the serovar. In addition, there was no significant difference between the MICs of the clinical isolates of the two treatment groups. After amplification by the PCR, chlamydial D N A was detected by both gel electrophoresis and spot hybridization in all 30 available samples t a k e n before the start of the treatment. Two of the 30 samples taken one week after the start o f the therapy were D N A positive o n gel electrophoresis. Both patients received azithromycin. Spot hybridization detected two additional D N A positive samples: one from a patient in the azithromycin group and one from a patient in the doxycycline group. O n e sample from a patient in the doxycycline group was not available for D N A detection. T h e first and third samples from this patient were both D N A positive. T h e o t h e r patient with a D N A positive sample taken four weeks after the start of the therapy received azithromycin. Table 2 shows the details of the D N A detection results. T h e serovar distribution
Table 1: Minima~inhibit~rycon~entrati~nsf~ra~3~is~ates~fCh~amydiatrach~matisandf~rthef~urser~vars. Antibiotic
MIC (gg/ml) Range (n = 31)
Polymyxin Ofloxaein Tetracycline Azithromycin Doxycycline Minocycline Roxithrornycin
40- > 160 0.5 - 1.0 0.03- 0.06 0.01 - 0.06 0.01- 0.02 0.01-0.03 0.01 - 0.03
Mean MIC (pg/ml) for serovar
Mean (n = 31)
Type D (n = 12)
Type E (n = t0)
Type F (n = 7)
Type J (n = 2)
NC* 0.516 0.034 0.029 0.019 0.017 0.017
NC* 0.500 0.030 0.030 0.019 0.018 0.015
NC* 0.500 0.033 0.030 0.020 0.019 0.018
NC* 0.571 0.030 0.031 0.019 0.013 0.016
NC* 0.500 0.060 0.015 0.020 0.020 0.020
*NC = Not calculated as MICs exceeded 160 ~ag/mt. Table 2: Numbers of positive results in culture and PCR for detection of Chlamydiatrachomatis.
Sample collection
Culture
Before treatment
PCR Electrophoresis
Spot hybridization
31/31 (100 %)
30/30~ (100 %)
30/30a (100 %)
One week after treatment
0/31
(0 %)
2/30b
(7 %)
4/30b (13 %)
Four weeks after treatment
0/29¢
(0 %)
2/29¢
(7 %)
2/29c
(7 %)
aThe first sample of one patient was not available for PCR; the second and third sample of this bPatient were negative by PCR. The second sample of one patient was not available for PCR; the first and third sample of this patient were positive by gel electrophoresis and spot hybridization. c No third samples were available from two patients for culture or PCR.
696
of the isolates of the patients with a D N A positive sample one or four weeks after therapy were: two isolates of serovar D, two isolates of serovar E and one isolate of serovar J.
Discussion
Samples were obtained from 60 patients in whom a previous sample had reacted positively in an enzyme immunoassay for detection of Chlamydia trachomatis antigen at a local laboratory. However, Chlamydia trachomatis was isolated from only 31 of these patients. During the study period the local laboratory performing the assays had experienced some technical difficulties resulting in a number of false positive results. This explains the apparently low recovery rate of our cultures. The problem of reduced performance of chlamydia detection assays in non-research laboratories has also been experienced in other countries (4). Using these 31 isolates the MIC values of seven antibiotics were determined. The method used is a modification of the method outlined by Ehret and Judson (5). Since the growth of many strains is limited in Eagle's minimal essential medium (6) we used Iscove's modified Dulbecco's medium. We did not include cycloheximide in the medium since it interferes with the production of normal infectious elementary bodies (7) and might therefore interfere with the action of the antibiotics. Since we were interested in obtaining the MIC values only we used microtiter plates for convenience. They were centrifuged at the same speed as the fiat bottom tubes and usually yielded the same number of inclusions per microscopic field as cover slips in flat bottom tubes. Polymyxin, included as a negative control, showed almost no activity against any of the clinical isolates. Ofloxacin, which so far has been shown to be the most active member of the fluoroquinolones (8), had an MIC of 0.5/ag/ml. Some studies indeed suggest this drug would have clinical efficacy (9). The other five drugs had MIC values that were very similar to each other. The MICs were within the range found previously (5, 10). The negative cultures after therapy showed that one week after therapy with a single dose of azithromycin or a seven-day course of doxycycline no viable Chlamydia trachomatis organisms were present at the sites of infection. Therefore both therapies eradicated live
Eur. J. Clin. Microbiol. Infect. Dis.
Chlamydia trachomatis organisms from these sites. However, chlamydial D N A could be detected by PCR in about 15 % (95 % confidence interval: 3 %-29 %) of the patients one week after the start of the therapy and in about 6 % (95 % confidence interval: 0 %-16 %) of the patients four weeks after therapy. The observation that four weeks after the start of therapy less samples were positive than after one week indicates that the presence of chlamydial D N A was a remnant of the infection rather than re-exposure. This is consistent with the fact that all patients stated they refrained from sexual intercourse until the second sample was taken for culture and with the negative cultures. In contrast with our finding Claas et al. (11) found no residual chlamydial D N A by the P C R one week after doxycycline therapy in 32 patients. However, using an antigen detection enzyme immunoassay Radcliffe et al. (12) found chlamydial antigen in 3 of 91 samples (95 % confidence interval: 0 % 7 %) one week after therapy. They concluded that this kind of method should not be used as a test of cure one week after therapy. Since the P C R is potentially much more sensitive than any antigen detection assay currently available and cannot discriminate between viable and non-viable microorganisms, we conclude that D N A detection by P C R is not suitable as a test of cure and cannot be used to monitor chlamydial infection. Further long term follow-up studies are needed to assess the value of the P C R in discriminating between D N A detected after a cured infection, in chronic infections and in acute (re)infections. Thus culture is still the only way to detect the presence of viable microorganisms and therefore the only way to study treatment failure or resistant strains (13). To detect possible differences in the MICs of serovars, all isolates were typed using monoclonal antibodies. The distribution of the serovars was similar to that found previously in the Netherlands (14), indicating that our study was not biased by an abnormal serovar distribution caused by selection of fast growing isolates or a local outbreak of one serovar. No difference was observed between MIC values of the different serovars. We found no differences between the two treatment groups with respect to the culture results, the MIC values, D N A detection and serovar types, and conclude that a single dose of azithromycin and a seven-day course of doxycycline are equally efficacious.
Vo1.11,1992
References 1. Foulds G, Shepard RM, Johnson RB: The pharmacokinetics of azithromyein in human serum and tissues. Journal of Antimicrobial Chemotherapy 1990, 25, Supplement A: 73-82. 2. Barnes RC, Wang SP, Kuo CC, Stamm WE: Rapid immunotyping of Chlamydia trachornatis with monoelonal antibodies in a solid phase enzyme immunoassay. Journal of Clinical Microbiology 1985, 22: 609613. 3. Sriprakash KS, Maeavoy ES: Characterization and sequence of a plasmid from the trachoma biovar of Chlamydia trachomatis. Plasmid 1987, 18: 205-214. 4. Taylor-Robinson D, Thomas BJ: Laboratory techniques for the diagnosis of chlamydial infections. Genitourinary Medicine 1991, 67: 256--266. 5. Ehret JM, Judson FN: Susceptibility testing of Chlamydia trachornatis: from eggs to monoclonal antibodies. Antimicrobial Agents and Chemotherapy 1988, 32: 1295-1299. 6. Johnson FWA, Hobson D: Factors affecting the sensitivity of replicating McCoy cells in the isolation and growth of Chlamydia A (TRIC agents). Journal of Hygiene 1976, 76: 441-451. 7. Karayiannis P, Hobson D, Lee N: Effect of cycloheximide on the infective yield of a genital strain of Chlarnydia trachomatis in McCoy cells. Infection and Immunity 1981, 33: 309-311.
697
8. Endtz HPH, Ossewaarde JM, van der Korput PJGM, Weiland HI':. In vitro activity of eight quinolones against Chlamydia trachomatis. Reviews of Infectious Diseases 1989, 11, Supplement 5: 1278. 9. Ronald AR, Peeling RW: Chlamydial infections and the quinolones. European Journal of Clinical Microbiology and Infectious Diseases 1991, 10: 351-354. 10. Seieux C, Bianchi A, Chappey B, Vassias I, Perol Y: In vitro activity of azithromycin against Chtamydia trachomatis. Journal of Antimierobial Chemotherapy 1990, 25, Supplement A: 7-10. 11. Claas HCJ, Wagenvoort JHT, Niesters HGM, "13oTT, van Rijsoort-Vos JH, Quint WGV: Diagnostic value of the polymerase chain reaction for ehlamydia detection as determined in a follow-up study. Journal of Clinical Microbiology 1991, 29: 42-45. 12. Radcliffe KW, Rowen D, Mereey DE, Mumlaz G, Ridgway GL, Robinson A J, Bingham JS: Is a test of cure necessary following treatment for cervical infection with Chlamydia trachomatis. Genitourinary Medicine 1990, 66:44 A,~..46. 13. Jones RB, van der Pol B, Marlin DH, Shepard MK: Partial characterization of Chlamydia trachomatis isolates resistant to multiple antibiotics, Journal of Infectious Diseases 1990, 162: 1309-1315. 14. Wagenvoort JIIT, Suchland RJ, Stamm WE: Serovar distribution of urogenital Chlamydia trachomatis strains in the Netherlands. Genitourinary Medicine 1988, 64: 159-161.
