Materials and Methods

Monkeys Forty-five sexually mature female pig-tailed macaques (Macaca nemestrina) were used in the study. All monkeys were housed at the Washington National Primate Research Center. Prior approval for use of monkeys in this protocol was obtained from the Institutional Animal Care and Use Committee at the University of Washington. Monkeys were handled humanely, and experiments were performed within the National Institutes of Health’s animal use guidelines

Inoculation Before inoculation, a baseline cervical examination and a minilaparotomy were performed, to assess cervical and tubal tissues and to obtain swabs for culture and ligase chain reaction (LCR) from both the lower and upper reproductive tract. Then, weekly for 5 weeks (figure 1), the cervix was inoculated by cervical washes of a serovar E C. trachomatis strain at a concentration of 5×10⁴ IFU/mL/inoculation. The strain was originally isolated from the endometrium of a patient with PID. The stock inoculum was divided into aliquots of 1×10⁵ IFU/mL in sucrose phosphate glutamate buffer and stored at −70°C until the day of use, when it was diluted to the inoculation titer

Treatments At the start of the study, each monkey was randomly assigned to 1 of 3 treatment groups: placebo (n=21), doxycyline (n=12; 2.2 mg/kg/day orally for 2 weeks), or azithromycin (n=12; 14 mg/kg/day orally for 1 week). Dose regimens were calculated to be analogous to US Centers for Disease Control and Prevention recommendations for the clinical use of these drugs for treatment of upper reproductive-tract C. trachomatis infection in humans. Treatments were initiated 1 week after the last C. trachomatis cervical challenge. There were more monkeys in the placebo group than in the treatment groups as part of a larger study objective in which we are assessing the influence of genetics on the progression of disease (data not shown)

Visual assessment Cervical examination was performed weekly throughout the study before each cervical inoculation and throughout the course of treatments. Minilaparotomy to assess the upper reproductive tract was done immediately before the first chlamydial inoculation and 1 week after the fifth cervical C. trachomatis challenge (1 day before the initiation of treatment). Final tissue samples were collected and assessed at the time of hysterectomy. At each visit, tissue erythema, edema, friability, and adhesions were noted by an observer blinded to treatment group. The findings at each upper reproductive-tract examination were documented by use of videography during each surgical procedure, and fallopian tubes were measured (at the distal ampulla) with a 10-mm ruler. The level of tubal edema was determined by comparing pretreatment and hysterectomy measurements with those recorded at baseline

Specimen collection To document C. trachomatis infection specimens were collected weekly throughout the experiment. Swabs were obtained from the cervix for culture and LCR. Additionally, at each minilaparotomy, swabs were collected from the fimbrial os for culture and LCR. Tissues collected at hysterectomy were fixed in 10% buffered formalin, embedded in paraffin, serially sectioned, and stained for tissue examination by light microscopy

Cell culture Specimens were cultured on cycloheximide-treated McCoy cells in 24-well microtiter plates with a single blind pass [18] and stained with monoclonal antibody (MAb) specific for chlamydia lipopolysaccharide. The cultures were inoculated both at their original concentration and at 10% dilution, to reduce the inhibition of successful culture. The number of inclusions was counted on the first cell passage

LCR assay Specimens were vortexed, incubated at 96°C–100°C for 15 min, and cooled to 20°C. The specimens were then either frozen for batched runs or refrigerated to be processed within 48 h in the LCR assay [19]. Specimens were subsequently added to the chlamydial LCR unit dose tubes (provided by Abbott Laboratories) and then amplified and processed in accordance with the manufacturer’s instructions

Hysterectomy A total hysterectomy was performed 4 weeks after the initiation of treatment (2 weeks after completion of doxycycline treatment and 3 weeks after completion of azithromycin treatment). The tissues of the reproductive tract were processed for examination by routine light microscopy and for immunopathologic and immunocytochemical assessment by use of antibodies to the chlamydial major outer membrane protein (MOMP) and chlamydia heat-shock protein 60 (CHSP60). Swabs were collected from the cervix and fallopian tubes for culture viability and LCR (the presence of DNA). All laboratory tests were performed in a blinded fashion, without knowledge of the clinical status of the monkeys, treatment groups, or test outcomes

Histologic and immunocytochemical analysis Tissues from the upper and lower reproductive tract of each monkey were stained for routine histologic assessment by use of hematoxylin-eosin. Each tissue section was assessed under 400× magnification. Plasma cells, polymorphonuclear cells (PMNs), and lymphocytes were morphologically identified and quantified by averaging their numbers in 5 nonadjacent fields

Parallel sections from paraffin-embedded tissues were incubated with C. trachomatis–specific MAb KK12, which recognizes the 40-kDa MOMP, or with antibody to CHSP60, which recognizes the 57-kDa genus-specific CHSP60 [20] (provided by Dr. Daniel Rockey, Oregon State University, Corvallis)

As an indicator of fibrosis, the presence of transforming growth factor (TGF)–β was quantified by staining with mouse anti–human TGF-β MAb TB21, which recognizes platelet-derived TGF-β (Serotec). The presence of TGF-β was quantified by recording the percentage of TGF-β–positive staining cells in 5 nonadjacent high-power fields. Normal mouse ascites fluid was used as a negative control. Peroxidase staining was done by use of the ABC Vectastain kit (Vector Laboratories), in accordance with the manufacturer’s instructions

Statistical analysis Statistical analysis of inflammatory infiltrates and the presence of specific antigenic proteins and TGF-β was first done by use of the Kruskal-Wallis&rank sum test, to identify differences between treatment groups. If a significant difference existed between groups, the Wilcoxon&rank sum test was used to test for pairwise differences between treatment groups

Results

Establishment of acute chlamydial infection Each monkey was inoculated in the cervix weekly for 5 weeks. Before treatment, all 45 inoculated monkeys were culture and/or LCR positive in the cervix. Before treatment, direct observation of the upper reproductive tract by minilaparotomy documented peritubal adhesion formation in 35 (78%) of 45 monkeys, fimbrial edema in 38 (84%), and tubal dilatation (measurement of tubal edema vs. baseline measurement) in 23 (51%)

Visual assessment of tissues at hysterectomy Tubal dilatation, erythema, and the presence and degree of adhesions were noted for each monkey at baseline, before treatment, and at the time of hysterectomy. As a result of treatment randomization, tissue profiles varied somewhat by treatment arm before the initiation of treatment. Therefore, to assess the effects of antimicrobial therapy, we analyzed whether findings progressed, resolved, or remained the same after the completion of treatment in each monkey

Overall, infection produced very few gross changes in the cervix. Cervical erythema and vesicles were noted only in monkeys in the placebo arm of the study. Minimal effects in either treatment arm were noted by gross observations of upper and lower reproductive-tract tissues (table 1). The progression of fimbrial erythema and edema, both of which were determined by use of somewhat subjective measures, varied only slightly across treatment groups. Friability was noted in 10% of the monkeys that received placebo before and after treatment, but its incidence increased from 8% to 33% in doxycycline-treated monkeys and resolved (from 17% before treatment to 0% at the time of hysterectomy) in azithromycin-treated monkeys. The occurrence of tubal erythema, dilatation, and adhesions increased in the placebo group. The incidence of tubal erythema resolved somewhat after doxycycline (from 50% to 25%) and azithromycin (from 25% to 8%) treatment. The incidence of tubal dilatation remained the same after doxycycline treatment (58%) and improved somewhat (from 67% to 50%) after azithromycin treatment. The number of peritubal adhesions increased after doxycycline treatment (from 75% to 83%) but remained constant (92%) after azithromycin treatment

Detection of C. trachomatis by culture and LCR from cervical and tubal tissues after treatment Of the 21 monkeys that received placebo (and therefore had essentially untreated chlamydial infection for 5–9 weeks before undergoing hysterectomy), 3 tested positive for chlamydial infection in the upper reproductive tract (fimbria) by LCR. Two of these monkeys also tested positive for cervical infection. A total of 15 placebo-treated monkeys tested positive for cervical chlamydial infection at the time of hysterectomy—12 by culture and LCR, and 3 by LCR alone (table 2)

By contrast, only 2 monkeys (of 12) in the doxycycline-treated group tested positive for chlamydial infection in the upper (n=1 culture positive) or lower (n=1 LCR positive) reproductive tract. None of the monkeys in the azithromycin-treated group tested positive for chlamydial infection at any site by either test

Detection of chlamydial MOMP and CHSP60 in cervical and tubal tissues after treatment Tissues collected at the time of hysterectomy were also assessed by immunocytochemical staining for chlamydia-specific MOMP and CHSP60. Of the 21 monkeys in the placebo treatment arm of the study, MOMP was detected in the upper reproductive tract of 8 monkeys and in the lower reproductive tract of 17. A subset of these monkeys also had CHSP60 detected in reproductive-tract tissues (2/8 in upper and 10/17 in lower). After 12 monkeys completed doxycycline treatment, MOMP was detected in lower reproductive-tract tissues of 4, 1 of which also tested positive for CHSP60 in the same tissues. MOMP and CHSP60 were not detected in any tissues of the 12 monkeys that received azithromycin treatment (table 2)

Effects of treatment on inflammatory infiltrates in the cervix and upper reproductive-tract tissues The tissues of monkeys randomized into the placebo treatment group had an inflammatory infiltrate typical of acute chlamydial infection in both the lower and upper reproductive tract. Specifically, an increase in plasma cells and PMNs, with a greater influx of lymphocytes, was noted in lower and upper reproductive-tract tissues collected at the time of hysterectomy (table 3). Monkeys treated with doxycycline developed inflammatory-cell profiles that were very similar to those of untreated monkeys. However, after azithromycin treatment, levels of inflammatory infiltrates were shown to be significantly reduced. Azithromycin treatment ameliorated the immune response to chlamydial infection, compared with placebo treatment, as evidenced by a significantly decreased presence of inflammatory cells. In contrast, doxycycline treatment did not suppress the inflammatory-cell infiltrate (table 3). The fimbria tissues of azithromycin-treated monkeys were characterized by fewer plasma cells, PMNs, and lymphocytes, compared with those of placebo-treated monkeys. The presence of plasma cells was significantly lower than that of doxycycline-treated monkeys as well. The cellular infiltrates in fallopian tubes looked similar across all 3 treatment groups. The endocervical and cervical tissues of azithromycin-treated monkeys each had significantly fewer plasma cells and PMNs than did those of placebo-treated monkeys. Again, levels of plasma cell infiltrates in the cervix were significantly less pronounced in azithromycin-treated monkeys than in doxycycline-treated monkeys. The only significant difference noted after doxycycline treatment, compared with placebo treatment, was a decreased influx of PMNs in cervical tissue. However, azithromycin-treated monkeys displayed even fewer cervical PMNs than did doxycycline-treated monkeys. Although doxycycline treatment modestly reduced the concentration of TGF-β in the cervix, endocervix, and fimbrial tissues, compared with placebo treatment, there was a much more marked and statistically significant reduction in levels of TGF-β in azithromycin-treated monkeys, compared with placebo-treated monkeys. This may indicate less development of fibrosis in the azithromycin-treated monkeys

Discussion

Despite the completion of appropriate antimicrobial treatment for chlamydial infection, some women develop PID, ectopic pregnancy, or tubal-factor infertility. It is not clear whether these sequelae result from reinfection, persistence of the original chlamydial infection, delayed treatment, lack of patient compliance with the treatment regimen, or antimicrobial resistance or whether they are associated with C. trachomatis at all. Despite the importance of this issue, the efficacy of antimicrobial treatment on upper reproductive-tract disease associated with chlamydial PID cannot be readily studied in humans. We thus used the macaque model of salpingitis to evaluate the effects of doxycycline and azithromycin on the treatment of genital chlamydial infection

The study reported here was designed in part to assess whether prolonged azithromycin treatment (1 week) would improve clearance of the chlamydial organism and antigen after multiple inoculations, when the disease is likely to have progressed to the upper reproductive-tract tissues. As measurements of treatment efficacy, we assessed lower and upper reproductive-tract secretions and tissues for the presence of chlamydial antigens and nucleic acids. The inflammatory reaction seen in chlamydia-infected upper reproductive-tract tissues was also monitored throughout the study. The data indicate that, in this monkey model, azithromycin more effectively prevented the progression of inflammation and fibrosis in both the lower and upper reproductive tract after repeated chlamydial infections

Eradication of the organism In the present study, azithromycin treatment completely eradicated chlamydial organisms from the upper and lower reproductive-tract tissues as measured by culture, LCR, and immunocytochemical detection of MOMP or CHSP60. After doxycycline treatment, only 1 monkey tested positive for chlamydia by cervical LCR, but 1 additional monkey tested positive by culture of a fimbrial swab collected at the time of hysterectomy; this same monkey, plus 3 others, had evidence of the chlamydia organism detected in cervical tissues by immunocytochemical staining. Overall, doxycycline treatment was thus less effective in clearing chlamydia from the reproductive tract than was azithromycin. As was expected, the majority of placebo-treated monkeys had continued evidence of the chlamydial organism detected throughout their reproductive-tract tissues until the time of hysterectomy. As in previous studies, despite the use of a highly sensitive culture system, we found it difficult to recover viable chlamydia from fimbrial or fallopian-tube specimens, despite positive tests for chlamydial antigen or genes from the same specimens. Correlation of culture with these other tests was much higher in cervical specimens, for which rates of culture positivity were high

Suppression of the inflammatory reaction to chlamydial infection According to histologic assessment of tissues collected at hysterectomy, doxycycline performed somewhat better than placebo treatment, primarily in cervical tissues, in reducing the inflammatory response. Azithromycin treatment, on the other hand, performed significantly better than placebo in almost all parameters assayed and in all tissues except the uterus. Significantly fewer plasma cells, PMNs, and lymphocytes were detected in the monkeys treated with azithromycin. In each tissue site throughout the female reproductive tract, evidence of TGF-β, a cytokine indicative of fibrosis, was significantly less prevalent in azithromycin-treated monkeys than in placebo-treated monkeys. These monkeys also had significantly less TGF-β identified in endocervical and fimbrial tissues than did doxycycline-treated monkeys. Azithromycin has been reported to inhibit the production of proinflammatory cytokines, such as interleukin (IL)–1β, IL-6, IL-8, and tumor necrosis factor–α [21 –24]. Although the mechanism of activity is not well understood, azithromycin may provide anti-inflammatory as well as antimicrobial properties. This combined activity may be particularly well suited for the treatment of chlamydial infection, particularly in the upper reproductive tract

Gross observations of upper reproductive-tract tissues indicated that azithromycin treatment may have arrested the progression of deleterious symptoms, in that evidence of friable tissue was diminished and occurrences of tubal edema decreased after azithromycin treatment. Potential effects of azithromycin treatment on gross pathologic results are difficult to interpret, because the pretreatment incidence of peritubal adhesions was high (11/12) in azithromycin-treated monkeys. Such adhesions have been implicated in tubal-factor infertility, which is a known sequelae of chlamydial infection. In contrast, placebo and doxycycline treatment did not affect ongoing fibrosis

In conclusion, azithromycin as a single 1-g dose is currently recommended for the treatment of uncomplicated lower reproductive-tract infection with C. trachomatis. The findings of the present study suggest that azithromycin administered for a longer period (7 days) should be considered for further study in upper reproductive-tract infection (PID) caused by C. trachomatis as well. Controlled studies that assess the role played by azithromycin in the treatment of human chlamydial infection of the upper reproductive tract are needed, including assessment of the drug’s effects on the inflammatory response