Abstract
Pelvic inflammatory disease (PID), the infection and inflammation of the female upper genital tract, is a common cause of infertility, chronic pain and ectopic pregnancy. Diagnosis and management are challenging, largely resulting from varying signs and symptoms and a polymicrobial etiology that is not fully delineated. Owing to the potential for serious sequelae, a low threshold for diagnosis and treatment is recommended. As PID has a multimicrobial etiology, including Neisseria gonorrhoeae, Chlamydial trachomatis and anaerobic and mycoplasmal bacteria, treatment of PID should consist of a broad spectrum antiobiotic regimen. Recent treatment trials have focused on shorter duration regimens, such as azithromycin, and monotherapies including ofloxacin, but data are sparse. Research comparing sequelae development by differing antimicrobial regimens is extremely limited, but will ultimately shape future treatment guidelines.
Keywords
Pelvic inflammatory disease (PID), the infection and inflammation of a woman's upper genital tract, occurs frequently among women of childbearing age. Estimates suggest that approximately 8% of US women and 15% of Swedish women are diagnosed with PID in their lifetime, with over 1 million women treated in the USA annually [1–3,101]. Rates of PID are much higher in developing countries and have been reported to be as high as 32% among Indigenous women in a remote community in the northern territory of Australia [4].
Pelvic inflammatory disease is thought to occur as microorganisms, frequently Chlamydial trachomatis and Neisseria gonorrhoeae, which ascend from the lower genital tract infecting the uterus, fallopian tubes and ovaries [5]. However, PID has a polymicrobial etiology, and up to 70% of cases are nongonococcal and nonchlamydial. Anaerobic Gram-negative rods, Mycoplasma genitalium and bacterial vaginosis (BV) are also associated with PID [6–10]. A cluster of BV-associated organisms (absence of hydrogen peroxide-producing lactobacillus, Gardnerella vaginalis, Mycoplasma hominis, anaerobic Gram-negative rods and, to a lesser degree, Ureaplasma urealyticum) has been associated with a twofold risk of incident PID (adjusted rate ratio: 2.03; 95% CI: 1.16, 3.53) [11]. As a direct result from damage to the cilia lining, the fallopian tubes, fallopian tube blockage or closure or adhesion formation, major reproductive and gynecologic sequelae, including infertility, ectopic pregnancy, recurrent PID and chronic pelvic pain, frequently follow PID [12,13]. Given the frequent and morbid nature of PID and its sequelae, the determination of ideal management strategies is important.
Pelvic inflammatory disease diagnosis
Both rapid and sensitive diagnosis and effective treatment are needed to optimally manage PID. Unfortunately, the diagnosis of PID is challenging and imprecise, as the symptoms and signs vary widely. Although pelvic pain is the most common symptom of PID, pelvic pain may be mild or even absent in some women, despite evidence of endometrial infection and inflammation [14,15]. Furthermore, PID may be confused with other pelvic conditions that exhibit similar symptoms, such as endometriosis, appendicitis or ectopic pregnancy. Owing to the serious reproductive morbidity associated with PID [12,13,16–20], a low threshold for both diagnosis and treatment is recommended. The following minimum criteria have been proposed by the US CDC: uterine/adnexal tenderness or cervical motion tenderness [21]. In order to increase sensitivity, not all criteria need to be present for diagnosis. Taken individually, cervical motion tenderness, uterine tenderness and adnexal tenderness have very high sensitivities compared with histologically confirmed endometritis (92, 94 and 96%, respectively) [22]. Requiring all three criteria results in a much lower sensitivity of 83% [22]. Similar to the low diagnostic threshold recommended by the CDC, the WHO suggests a syndromic management approach for the diagnosis of PID, in which patients presenting with lower abdominal pain, who are found on pelvic examination to have cervical motion tenderness, lower abdominal tenderness or uterine or adnexal tenderness, are treated immediately, without requiring additional potentially time-consuming and costly laboratory tests [102].
This immediate management strategy is supported by studies demonstrating relationships between delayed PID treatment and both infertility and chronic pelvic pain [23–26]. Additional CDC recommended diagnostic criteria to increase the specificity of PID diagnosis and identify other possible causes of pelvic pain include an oral temperature greater than 101°F (38°C), abnormal cervical or vaginal mucopurulent discharge, presence of white blood cells (WBCs) on saline microscopy of vaginal secretions, elevated erythrocyte sedimentation rate, elevated C-reactive protein and laboratory documentation of cervical infection with N. gonorrhoeae or C. trachomatis. However, in a recent analysis from the landmark Scandinavian Lund study of PID, none of the following variables had both high specificity and sensitivity for laparoscopically verified PID: abnormal vaginal discharge, fever above 38°C, vomiting, menstrual irregularity, ongoing bleeding, symptoms of urethritis, rectal temperature above 38°C, marked tenderness of pelvic organs on bimanual examination, adnexal mass and erythrocyte sedimentation rate greater than or equal to 15 mm in the first hour [27]. In fact, most had low specificity and sensitivity. Tenderness of pelvic organs did have a high sensitivity (99%) but an extremely poor specificity (<1%). In a study of 121 women meeting the CDC's minimal criteria for PID, vaginal WBC count was found to be a sensitive marker of upper genital tract infection [28]. Additional work is needed to verify the diagnostic accuracy of vaginal WBC count as well as other inflammatory markers, such as cytokines and interferons. Clearly, diagnostic criteria, largely based on empirical data, would benefit from additional evaluation and revision.
Pelvic inflammatory disease can be confirmed using laparoscopic or histopathologic examination of endometrial or fallopian tube biopsies, with laparoscopy traditionally considered the gold standard for PID diagnosis. However, neither of these confirmatory methods is extremely precise. Compared with laparoscopically diagnosed salpingitis, histologically confirmed endometritis has a sensitivity of 70–89% and a specificity of 67–92% [29–31]. Although considered the gold standard, laparoscopy itself has an extremely low sensitivity for PID diagnosis, ranging from 25 to 50% when compared with histopathologic evidence of PID demonstrated by fimbrial minibiopsy [32,33]. Furthermore, laparoscopy, which lacks standardization and relies on subjective interpretation of pelvic structure photographs, has only a fair intraobserver reproducibility for the diagnosis of PID (º = 0.58) and a poor-to-fair inter-observer reproducibility (º = 0.43) [32]. In addition to concerns about its sensitivity and standardization, laparoscopy is an invasive procedure that is not routinely used in clinical practice. MRI may prove to be a valid alternative diagnostic procedure, as limited data suggest high sensitivity (95%) and specificity (89%) for the diagnosis of PID, compared with laparoscopy [34]. Although MRI facilities are widely available in the USA and other developed countries, high cost and lack of availability in resource-poor settings may limit its broad diagnostic utility. Transvaginal ultrasound is another minimally invasive procedure that has been proposed to improve the diagnosis of PID, although it has a much lower sensitivity for laparoscopically diagnosed PID, ranging from 32 to 81% [34,35]. Power Doppler, a recent innovation that allows improved detection of blood flow and inflammation-induced hyperemia, has been found to have both a high sensitivity (100%) and specificity (80%), compared with laparoscopically confirmed PID [36].
Treatment regimens
Among women with mild-to-moderate PID, there is no difference in short-term clinical and microbiologic improvement or subsequent rates of infertility. In addition, there is no difference in recurrent PID, chronic pelvic pain or ectopic pregnancy between women hospitalized for PID and those treated as outpatients [37], even among younger women and those presenting with more severe disease, as measured by elevated temperature, WBC count or pelvic tenderness [38]. This suggests that with few exceptions, the vast majority of women with mild-to-moderate PID can be treated as outpatients with antimicrobial therapy, which is substantially less costly than hospitalization. The CDC does recommend hospitalization for certain subgroups of patients, including women for whom competing surgical emergencies such as appendicitis can not be ruled out, pregnant women, those who do not respond well to outpatient antibiotic therapy, patients who are unable to tolerate an outpatient oral regimen, women with severe illness, nausea and vomiting or a high fever and patients with tubo-ovarian abscesses [21]. Owing to its polymicrobial nature, PID is treated with antibiotics covering a broad spectrum of pathogens. CDC guidelines recommend outpatient treatment of PID with ofloxacin, levofloxacin, ceftriaxone plus doxycycline or cefoxitin and probenecid plus doxycycline; all with optional metronidazole for full coverage against anaerobes and BV
2006 US CDC recommended outpatient regimens for the treatment of pelvic inflammatory disease.
Clindamycin and aminoglycoside (clinical cure 92% and microbiologic cure 97%);
Cefoxitin and doxycycline (clinical cure 93% and microbiologic cure 98%);
Cefotetan and doxycycline (clinical cure 94% and microbiologic cure 100%);
Ciprofloxacin (clinical cure 94% and microbiologic cure 96%) [39].
A fifth inpatient regimen including metronidazole and doxycycline was found to have much lower clinical cure (75%) and microbiologic cure (71%) rates [39]. These low-efficacy rates are probably due to the poor coverage of this latter combination against N. gonorrhoeae [39]. One outpatient regimen (cefoxitin, probenecid and doxycycline) was included in this meta-analysis and was found to have a pooled clinical cure rate of 95% and a microbiologic cure rate of 91% [39]. Since the publication of this meta-analysis, additional studies of PID treatment have been conducted, including several new monotherapies.
Fluoroquinolone trials
Recent randomized clinical trials for the treatment of pelvic inflammatory disease.
b.i.d.: Twice daily; im.: Intramuscular; iv.: Intravenous; t.i.d.: Three-times daily.
Two nonrandomized studies of laparoscopically confirmed salpingitis similarly support the efficacy of ofloxacin for gonococcal and chlamydial PID by demonstrating that ofloxacin, administered every 12 h intravenously, followed by a 10–14-day oral regimen, was associated with a 100% gonococcal cure rate [41,42], a near 100% chlamydial cure rate [41,42] and a 98% clinical cure rate [42]. However, although all patients with anaerobic bacteria cultured at study admission were considered clinically cured at follow-up in one of these studies, follow-up anaerobic cultures were not reported and, thus, microbiologic cure of anaerobes can not be determined [42].
Since the lack of anaerobic coverage with ofloxacin is a concern, emphasized by the high rate of treatment failure among patients with nongonococcal, nonchlamydial PID [40], the CDC suggests the optional addition of metronidazole [21]. Alternatively, one randomized clinical trial of 131 women with laparoscopically confirmed PID investigated another fluoroquinolone, ciprofloxacin plus clindamycin regimen [43]. In this study, clindamycin/ciprofloxacin was found to be as effective for the clinical cure of PID (97 vs 95%, respectively) and C. trachomatis eradication (100% in both groups) as compared with ceftriaxone plus doxycycline [43]. N. gonorrhoeae was less prevalent, and although the microbiologic cure was only 50% among women treated with clindamycin plus ciprofloxacin, only two women tested positive at baseline. Although aerobic and anaerobic isolates were frequently identified before treatment, no comparisons of microbiologic cure for these pathogens were presented.
In the most recent fluoroquinolone randomized clinical trial by Ross et al., moxifloxacin was found to have high rates of clinical resolution (90%) and microbiologic cure [44]. Microbiologic cure was 100% for N. gonorrhoeae, Mycoplasma hominis, Mycoplasma genitalium, Escherichia coli and other Gram-negative anaerobes. Although C. trachomatis eradication was lower at 89%, it was slightly higher than that of the comparator regimen, ofloxacin plus metronidazole (86%) [44]. Furthermore, the N. gonorrhoeae eradication rate was much higher, (100 vs 82%) and drug-related adverse events were less (23 vs 31%) among women treated with moxifloxacin.
Azithromycin trials
As a result of the enhanced adherence and widespread use for treating chlamydial infections, a handful of studies have examined the efficacy of single- or short-duration dose azithromycin in PID treatment. Not surprising, adherence for single-dose azithromycin therapy has been reported as 100% [45]. In an Indian randomized clinical trial of 165 women with clinically suspected PID, a kit containing one tablet of fluconazole (150 mg), one tablet of azithromycin (1 g) and two tablets of secnidazole (2 g) was associated with a PID clinical cure rate of 94%, similar to that among a group treated with ciprofloxacin plus tinidazole for 7 days (96%) and better than that observed for women treated with doxycycline plus metronidazole for 1 week (91%) [46]. However, although this trial demonstrated high rates of clinical cure, data on microbiologic cure were not presented. In a UK randomized clinical trial comparing azithromycin monotherapy, azithromycin plus metronidazole and standard 21-day regimens of metronidazole/doxycycline/cefoxitin/probenecid or doxycycline/amoxicillin/clavulanate among 309 patients with PID, azithromycin was found to have a high rate of clinical success, similar to other regimens (97% azithromycin monotherapy vs 98% azithromycin/metronidazole vs 95% comparator regimen). Moreover, azithromycin provided excellent rates of C. trachomatis, N. gonorrhoeae, M. hominis and anaerobe eradication [47]. However, complicating the interpretation of this study was the combining of two different trials in the analyses. In one trial, azithromycin 500 mg was administered intravenously on day one followed by 6 days of oral azithromycin 500 mg. In the other trial, the intravenous protocol was conducted for 2 days followed by 5 days of oral azithromycin therapy. The comparator regimens also differed in each trial. Therefore, the optimal azithromycin regimen can not be determined. Furthermore, the dropout rate at the final follow-up was extremely high at 78%, and this reduces the validity and generalizing ability of the microbiologic cure evaluation. Most recently, in a randomized clinical trial among 106 patients first receiving an intramuscular injection of ceftriaxone 250 mg, those assigned to additionally receive azithromycin 1 g/week for 2 weeks experienced a higher rate of clinical cure compared with those whose regimen included doxycycline 200 mg/day for 2 weeks (98 vs 86%, respectively; p = 0.02) [48]. Although pathogens were not identified in the majority of patients and anaerobes were not evaluated, clinical cure for C. trachomatis (6/6, 100%) and M. genitalium (6/7, 86%) was high among the entire study cohort. Data on M. genitalium and T. vaginalis cures were not presented by this study arm, although the authors did report that the M. genitalium treatment-failure case as well as one additional T. vaginalis treatment-failure case occurred within the azithromycin group. This may be explained by the fact that although M. genitalium is susceptible to macrolides, azithromycin-resistant strains have recently been identified [49].
Doxycycline & metronidazole trials
A few studies have further examined the outpatient combination regimen of doxycycline and metronidazole. In a randomized clinical trial of 40 patients with laparoscopically confirmed salpingitis, combined doxycycline/metronidazole treatment exhibited a low clinical cure rate of 35%, with a 50% clinical improvement rate [50]. The low efficacy of this regimen was replicated in an observational UK study of 135 women with PID, in which only 55% experienced a clinical cure 2 or 4 weeks following treatment [51]. The addition of ceftriaxone to this regimen resulted in a higher but still suboptimal cure rate of 72% [51]. As combined doxycycline/metronidazole provides poor coverage against N. gonorrhoeae and was also associated with the lowest pooled clinical and microbiologic cure rates in the meta-analysis conducted by Walker and colleagues, this is not considered to be an optimal regimen for PID treatment.
Inpatient meropenem trial
Most randomized clinical trials of PID mono-therapy have been conducted among women with mild-to-moderate PID treated as outpatients. In one study of 84 women hospitalized for PID, meropenem, which has demonstrated in vitro activity against a broad spectrum of Gram-negative and Gram-positive aerobic and anaerobic bacteria, was found to have a high and relatively similar clinical response (88 vs 90%, respectively) and rate of microbiologic cure (88 vs 86%, respectively) as compared with treatment with clindamycin plus gentamicin [52]. However, only 37% of patients had cultures performed after treatment and were included in comparisons of microbiologic cure. Further studies are needed to confirm the use of this well-tolerated broad-spectrum inpatient monotherapy regimen for the treatment of PID [52].
Prevention of reproductive morbidity
Randomized clinical trials of PID treatment primarily define clinical cure by reduction of PID-related signs and symptoms. Although the end goal is to prevent long-term sequelae, evidence suggests that short-term markers of clinical cure, including tenderness at 5 and 30 days, are not predictive of PID-related reproductive morbidity [53]. There is limited data on the efficacy of any PID treatment regimen in the prevention of infertility or other adverse reproductive sequelae. From the PID Evaluation and Clinical Health (PEACH) study, we have previously reported that among women with clinically suspected mild-to-moderate PID treated with the standard antibiotic regimen of doxycycline and cefoxitin, endometritis and upper genital tract gonococcal and chlamydial infection were not associated with reproductive morbidity [54]. One interpretation is that the antibiotics used to treat modernday chlamydial and gonococcal PID may be so effective that there is no increase in future reproductive morbidity rates among women with treated endometritis. This elucidation is supported by the high conception rate (89%) reported among a Finish cohort of 39 women treated as inpatients with doxycycline and metronidazole for mild-to-severe salpingitis [55]. However, the average time to pregnancy in this cohort was 38 months, with longer times to pregnancy among women with severe salpingitis. It is possible that many of these women would have indeed been considered infertile at some point during follow-up, as infertility is often defined as a lack of conception after 12 months of unprotected intercourse. In addition, the high conception rate achieved over the 10-year follow-up may be partially attributed to successful infertility treatment among women with post-PID infertility. Furthermore, no studies have compared the rates of reproductive morbidity following treatment with modern-day PID regimens to the rates of reproductive morbidity among a control group without PID. Thus, it may be possible that all women treated for PID – whether gonococcal/chlamydial or nongonococcal/nonchlamydial – experience greater reproductive morbidity than women without PID, despite antibiotic treatment. Certainly, tubal scarring following antibiotic treatment for PID is common, with estimates ranging from 33 to 45% [56–58].
It is also possible that certain subgroups of women with PID may be more likely to experience adverse sequelae. In the PEACH study we have reported that compared with the lack of each respective microorganism, women with nongonococcal bacteria identified in the endometrium were generally more likely to suffer reproductive morbidity than women with endometrial gonococcal infection, despite the fact that all women in the PEACH study were treated with doxycycline and cefoxitin (infertility rates: N. gonorrhoeae = 13%, C. trachomatis = 19%, anaerobic bacteria = 22%, Ureaplasma urealyticum = 27% and M. hominis = 17%; chronic pelvic pain rates: N. gonorrhoeae = 27%, C. trachomatis = 21%, anaerobic bacteria = 33%, U. urealyticum = 41% and M. hominis = 54%). Similarly, in a study of 50 women with laparoscopically confirmed salpingitis by Brunham and colleagues, 54% of women with nongonococcal infections suffered future adverse reproductive outcomes compared with none of the women with gonococcal infections [59]. Collectively, the PEACH study and the study by Brunham and colleagues suggest that the standard antibiotic regimens of doxycycline/cefoxitin [54], doxycycline/clindamycin [59] and doxycycline/metronidazole [59] may not be optimal for the prevention of adverse sequelae among women with nongonococcal endometritis and salpingitis.
Animal models may provide insight into better regimens for nongonococcal PID treatment. Rapid single-dose oral azithromycin therapy has been found to prevent infertility in a mouse model of chlamydial salpingitis [60]. Similarly, azithromycin was found to be more effective than doxycycline in the microbiologic cure of C. trachomatis infection and the prevention of immunopathologic upper reproductive tract damage in the Macaque model of PID [61]. Studies of reproductive sequelae following various PID treatment regimens in humans are needed.
Patient & provider adherence
Short-term dose and monotherapies for PID may be particularly important for groups with reportedly low rates of adherence and access to care, including adolescents and lower socioeconomic status groups. Adolescents are at increased risk for STDs, and thus PID, as they are more likely to have unprotected intercourse, are more likely to have a greater number of current sexual partners, change partners more frequently and adolescent women are biologically susceptible to infection due to greater cervical ectopy. Since many adolescent women rely on emergency departments and outpatient clinics for the evaluation and treatment of STD symptoms, the need for a low diagnostic and management threshold for PID is even more critical among these young women, as the likelihood for additional follow-up care is low. Furthermore, as the risk of adverse sequelae, such as infertility and chronic pain, is increased if PID treatment is delayed [23–26], rapidly treating women of young reproductive age during their first medical visit is of utmost importance. Not surprisingly, data from the US National Hospital Ambulatory Medical Care Survey published in 2004 suggests an extremely low rate of full adherence (35%) with CDC PID treatment recommendations among women presenting to emergency departments [62]. An astounding 20% of patients with PID presenting to emergency departments received no treatment or treatment not in agreement with the guidelines [62], although azithromycin was one of the drugs prescribed among women not treated in compliance with the CDC PID treatment guidelines published at that time [62]. Similarly, a 2004 retrospective chart review of 121 PID cases presenting to an urban hospital emergency department revealed extremely low rates of accordance with the following CDC treatment guideline domains: history (14%), physical examination (22%), diagnostic testing (71%), antibiotic use (32%) and safe-sex instructions (15%) [63]. A common discrepancy with the CDC guidelines was the administration of single-dose azithromycin therapy (35% of patients) and less than the full 14 days of doxycycline therapy (22% of patients) [63]. Collectively, these studies suggest that emergency department clinicians may choose to treat women with PID using shorter duration antibiotics, with the goal of increased adherence. There are limited data to support the efficacy of short-duration regimens, underscoring the need for evaluation of this approach for PID treatment. However, both diagnosis and treatment is clearly suboptimal in the adolescent and emergency department populations. Data suggest that interventions utilizing algorithms and clinical practice guidelines based on CDC treatment recommendations can improve the quality of care provided to adolescents diagnosed with PID in emergency rooms and outpatient clinics [64]. However, whereas provider adherence to CDC treatment guidelines increased from 38 to 91% during this intervention study, approximately 40% of treated adolescents assessed at follow-up did not complete all doses of medication, suggesting that many adolescent women continue to have difficulty with medication adherence despite improvements in provider compliance [64]. Barriers to care and adherence, particularly among adolescent women, should be taken into consideration in the study and management of PID.
Management of subclinical pelvic inflammatory disease
Since women with PID present with varying signs and symptoms, the diagnosis and management of PID is not straightforward. In the case of silent PID, women have mild or no pelvic pain, despite evidence of endometritis or salpingitis [15]. C. trachomatis is associated with an excess of a threefold increase in subclinical PID risk (OR: 3.4; 95% CI: 1.8, 6.3), compared with a twofold risk of subclinical PID among women with N. gonorrhoeae (OR: 2.4; 95% CI: 1.1, 5.1) [14]. As rates of C. trachomatis continue to rise and N. gonorrhoeae rates decline [103], a predominance of chlamydial upper genital tract infection may correspond to an increase in silent PID.
In a study of 207 women presenting to the Seattle-King County STD Clinic without clinical evidence of PID, histologic endometritis was identified among 18% of the women. Women were treated with a single oral dose of cefixime 400 mg, a single oral dose of azithromycin 1 g and metronidazole 500 mg twice daily for 7 days. Among 18 women with endometritis and endometrial biopsies conducted both before and after therapy, endometritis was resolved in 89% of cases. However, whether treatment for subacute endometritis can prevent long-term PID sequelae is unknown. In addition, the identification of women with subacute endometritis to target for treatment remains a challenge, as laparoscopy and endometrial biopsy are not routinely used for the diagnosis of PID.
Conclusion
The diagnosis and management of PID remains a challenge. Although N. gonorrhoeae and C. trachomatis account for a sizeable portion of cases, the etiologic agent is unidentified in the majority of women with PID. Progress is being made in the identification of anaerobic causes of PID, and further research into novel PID pathogens will advance our understanding of PID pathophysiology and shape treatment trials and guidelines. Ultimately, microbe-specific and optimized treatment may preserve fertility following PID and also prevent recurrent and persistent infection, ectopic pregnancy and chronic pain, improving the long-term prognosis of women diagnosed with PID.
Current CDC guidelines recommend outpatient treatment of PID with ofloxacin, levofloxacin, ceftriaxone plus doxycycline or cefoxitin and probenecid plus doxycycline; all with or without the addition of metronidazole for full coverage against anaerobes and BV
Among the recent PID treatment trials, regimens including ofloxacin, moxifloxacin, azithromycin and clindamycin/ciprofloxacin all yielded high rates of clinical cure and N. gonorrhoeae and C. trachomatis eradication, although microbiologic cure data among women with anaerobic PID are limited. This is due in part to the complexity of anaerobic PID study, as anaerobes may be present as commensal bacteria in the lower genital tract, and transcervical sampling of the endometrium may result in contamination of endometrial biopsy specimens by vaginal or cervical microorganisms. However, we have previously demonstrated that contamination probably does not account for the relationships between BV-associated microorganisms and acute endometritis. In the PEACH study, we have shown that anaerobic Gram-negative bacteria are associated with acute endometritis independent of BV [7]. Furthermore, anaerobic Gram-negative rods and anaerobic Gram-positive cocci, but not other organisms frequent among women with BV (i.e., G. vaginalis and M. hominis), are associated with acute endometritis [7]. Another barrier to the study of PID microbiologic cure is the lack of a 100% correlation between endometrial and fallopian tube microbiology and the unlikely determination of fallopian tube microbiology post-treatment.
Future perspective
Despite the challenges facing the study of anaerobic PID, since only a third to a half of PID cases are attributed to N. gonorrhoeae and/or C. trachomatis, and because BV, anaerobic Gram-negative rods and mycoplasmal bacteria have been identified among women with PID [7–10,81], the microbiologic efficacy of PID treatment regimens for both chlamydial/gonococcal and nonchlamydial/nongonococcal PID should be determined. Over the next 5 years, several areas of study are likely to take center stage, focusing on the identification of novel pathogens associated with PID, so that existing regimens can be examined for microbiologic efficacy and new broad-spectrum regimens can be tested and proposed as needed. As the microbiologic etiology is unknown in a substantial number of women with PID, it is possible that current regimens do not provide optimal microbiologic cure for a full range of PID-associated pathogens. There are a number of emerging pathogens currently under study for their role in BV and PID. Most data are currently available on the pathogen M. genitalium, which has recently been implicated in PID [8,10,27,81] and tubal-factor infertility [82,83], lacks a cell wall and, therefore, is resistant to cell wall-inhibiting antibiotics, including cephalosporin. M. genitalium has been found to persist among men treated with levofloxacin [84,85] and tetracyclines [86–89] for nongonococcal urethritis. Thus, a number of currently recommended PID treatment regimens are probably insufficient for the treatment of M. genitalium upper genital tract infection. Investigations of regimens targeted toward novel PID pathogens are likely to direct modifications of existing PID treatment protocols, improving both microbiologic and clinical cure rates. M. genitalium has demonstrated variable resistance to fluroquinolones (e.g., ciprofloxacin) [90,91], and susceptibility to macrolides, although azithromycin-resistant strains have recently been identified [49]. A newer quinolone, moxiflocacin, has recently been shown to exhibit a high degree of activity against M. genitalium [92].
Since delayed PID treatment and recurrent PID are risk factors for subsequent infertility and chronic pelvic pain [23–26], the next several years of PID treatment research should examine regimens that will both increase adherence and minimize treatment resistance. A promising drug, which may address both of these issues, is azithromycin. Probably owing to its short dosing schedule and high rates of associated adherence, azithromycin is already frequently used for the treatment of PID in emergency departments [62,63]. However, only a few studies have examined azithromycin therapy for PID treatment [45–47]. Further study of clinical cure, microbiologic cure and reproductive sequelae following treatment of PID with azithromycin will probably be key components in PID research over the next several years.
Lastly, clinical trials comparing reproductive and gynecologic morbidity between PID treatment regimens are greatly needed. Whether currently prescribed PID antibiotic regimens are effective in the prevention of subsequent reproductive morbidity is largely unknown. Arguably, long-term PID sequelae represent the most important treatment outcomes. Ultimately, microbe-specific and optimized PID treatment should preserve fertility and prevent recurrent and persistent infection, ectopic pregnancy and chronic pain, improving the long-term prognosis of women diagnosed with PID.
Executive summary
Pelvic inflammatory disease (PID) is a frequent infection of the female upper genital tract commonly associated with infertility, chronic pelvic pain, ectopic pregnancy and recurrent infection.
Owing to the potential for serious sequelae, a low threshold including the following criteria should be used for the diagnosis of PID: uterine/adnexal tenderness or cervical motion tenderness.
Since PID has a polymicrobial etiology including Neisseria gonorrhoeae, Chlamydial trachomatis and anaerobic and mycoplasmal bacteria, treatment of PID should be broad spectrum.
Single or short-term dose and monotherapies for PID may increase treatment adherence, particularly among populations with limited access to care.
The identification of the role and treatment of novel pathogens in PID is needed.
Further study of short-duration PID regimens, including azithromycin, is desirable.
The comparison of reproductive and gynecologic morbidity between PID treatment regimens is greatly required.
Footnotes
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.