1. Introduction
Genital infection by the bacterial pathogen
Chlamydia trachomatis (
C.tr.) has been known for many years to function as a trigger for Reiter’s syndrome (RS) and reactive arthritis (ReA) [
1,
2]. Recent studies further indicate a chlamydial aetiology even for patients with spondyloarthritis (SpA) [
3,
4]. Epidemiologic data suggest that
Chlamydia-induced reactive arthritis (CReA) is a more common condition than previously thought, and that clinicians often fail to recognize it [
5,
6]. Most importantly, a recent controlled study demonstrated that CReA can be successfully treated with combination antibiotic therapy, thereby raising the possibility of a cure [
7,
8]. That possibility highlights the increased significance of awareness and diagnosis of arthritis and SpA elicited by
Chlamydiae.
It has become clear during the last two decades that two chlamydial species are responsible for causing arthritis,
C.tr. and the related respiratory pathogen
Chlamydia (Chlamydophila) pneumoniae (
C.pn.), and that both elicit the disease via their persistent presence in the joint [
1,
2,
9,
10].
C.pn. is essentially ubiquitous in all populations so far examined, raising the possibility that coinfection involving the two chlamydial species might be significant in the aetiology of ReA and/or uSpA. This possibility was reinforced by the observation that DNA from a wide variety of bacterial species can be found in the joints of patients with arthritis [
11].
Polymicrobial or coinfections are well-known and of demonstrated clinical importance in infectious diseases of the oral cavity, in otitis media, in diabetic foot wound infections, chronic infection in the cystic fibrosis lung, and in other clinical entities [
12]. In general, coinfections can be concurrent, as in bacterial pneumonia with
Staphylococcus aureus complicating flu infection, or they can be closely sequential, as with respiratory viruses plus commensal bacteria such as in otitis media caused by bacterial
Streptococcus pneumoniae or
Haemophilus influenzae following coronavirus, respiratory syncytial virus, or adenoviral infection [
13]. Importantly, in addition to such acute coinfections, chronic infections such as those involving human immunodeficiency virus (HIV) can lead to concurrent bacterial infections with
Mycobacterium tuberculosis and other pathogens [
13].
Coinfections involving
Chlamydia trachomatis (
C.tr.) and
Neisseria gonorrhoeae have been described often in screening programs and clinical settings; patients with gonorrhoeae also have been reported to have a concurrent chlamydial infection in less than 1% to a high of 70% of individuals, thus demonstrating wide variation depending on the population examined. [
14,
15,
16,
17,
18,
19]. Other coinfections of relevance in urogenital contexts include genital mycoplasmas and genital ureaplasmas. Coinfections with
Mycoplasma genitalium or
Ureaplasma urealyticum biovar 2 in men with gonococcal urethritis are associated with post-gonococcal urethritis, independent of
C.tr. [
20]. Additional studies in healthy individuals, in women in a cross-sectional sexual transmission infection (STI) screening program, and in non-gonococcal urethritis and chronic prostatitis, have reported coinfections of
C.tr. with
M. genitalium,
M. hominis,
U. urealyticum and
U. parvum in urogenital specimens [
21,
22,
23,
24,
25,
26,
27,
28,
29]. These have been implicated in sexually-transmitted urogenital diseases, although the evidence of the pathogenic role of
Ureaplasma species is questionable given its commensal state in the urogenital flora [
23,
24,
27,
28,
29].
As mentioned, compelling evidence has accumulated in recent years supporting the causative role of both
C.tr. and
C.pn. in ReA and spondyloarthritis (SpA) [
9]. In this article, we review the evidence for coinfections involving chlamydial species reported in patients with that arthritis and SpA. We discuss the potential aethiopathogenic and clinical implications of such infections, and we address the need for future basic research and clinical studies to improve diagnosis, clinical description, and treatment.
2. Coinfections with Chlamydiae in Reactive Arthritis and Spondyloarthritis
Coinfections involving
C.tr. and
C.pn. were first described in synovial tissue (ST) of patients with reactive arthritis (ReA), Reiter’s syndrome (RS), and rheumatoid arthritis (RA), and later also in patients with chronic ReA and undifferentiated spondyloarthritis (uSpA) [
3,
5,
7,
8,
9] Most recently, for the first time, multiple intra-articular coinfections of
Chlamydiae with
Mycoplasma and
Ureaplasma were reported in patients with post-venereal ReA [
30,
31]. As mentioned, during the last two decades it has been established that
C.tr. and
C.pn. also cause SpA in at least some patients, due to their ability to persist in the joint [
2,
3,
4]. The epidemiologic prevalence of infection with
C.pn. is significantly higher than that for
C.tr. overall, an observation which provided a reason for Schumacher and colleagues to assess the incidence of
C.pn. DNA compared to that of
C.tr. in ST and several synovial fluid (SF) samples from patients with ReA, other arthritides, and in normal joints [
9]. Only 12.7% of the samples (
n = 217) were positive for
C.pn. compared to 28.8% for
C.tr. Importantly, 2.4% were positive for both organisms; 5.3% of patients diagnosed with ReA; and 4.7% diagnosed with RA were PCR-positive for
C.pn. DNA. No clear differentiating clinical or other features were identified in the patients positive for both chlamydial species, which is the reason it was not possible to decide which of the two was the causative agent for disease in these cases. Interestingly, genital infection with
C.tr. is responsible for eliciting up to half of all cases of ReA, while pulmonary infection with
C.pn. is responsible for less than 15% of cases. The basis for this discrepancy is unknown, but must be related to details of the genetic component of each. However, the elicitation of significant levels of synovial inflammation by either
C.tr. or
C.pn. does appear to be accomplished by congruent means. That is, transcription of the highly proinflammatory chlamydial hsp60 protein is upregulated during persistent synovial infection in both of these chlamydial species [
32,
33].
Undifferentiated SpA has been suggested to be a forme fruste of ReA, based on indirect serological evidence of preceding genitourinary or enteric infection [
34]. Carter and colleagues investigated the prevalence of
C.tr. and/or
C.pn. DNA by PCR in ST and peripheral blood monocytes (PBMC) in patients with chronic uSpA (
n = 26), using patients with osteoarthritis (OA) (
n = 167) as controls [
3]. Thirty-eight percent of patients with uSpA were positive in ST and 38% for
C.tr., 15% for
C.pn., and 8% for both together; OA patients were 11%, 0%, and 0.6% positive, respectively. Only 2 patients with uSpA had a history of possible
C.tr. infection, and none had a history of
C.pn. infection. PBMC were positive for chlamydial DNA in only 4/26 (15%) patients with uSpA (3
C.tr.; 1
C.pn.); of those, 2 were positive in ST (1
C.tr.; 1
C.pn.). Together, these data suggested that chlamydial infections, which are often occult for both organisms, are aetiologic for many patients with uSpA. Interestingly, some patients with OA, a degenerative disease without presumed infective aetiology, were found also positive for chlamydial DNA, however, less frequently than those with ReA; this observation clearly suggests that some level of subclinical, essentially invisible, background infection is present in the populations examined.
The question of the role of
Chlamydiae as innocent bystander or causative agent in joint disease was investigated in a double-blind, placebo-controlled six-month trial with combination antibiotics (doxycycline 100 mg twice daily and rifampin 300 mg daily, azithromycin 500 mg daily × 5 days then twice weekly and rifampin 300 mg daily). In patients with chronic CReA, all were PCR-positive for
C.tr. or
C.pn. DNA in PBMCs and/or ST [
7]. Sixty-three percent of patients undergoing active treatment were responders compared to 20% under placebo. Six (22%) patients undergoing antibiotic treatment experienced complete remission, compared to none in the placebo arm. Most interestingly, 5/6 patients who went into remission were in the azithromycin and rifampin treatment arm, suggesting this combination is most effective. In this study, coinfections of
C.tr. plus
C.pn. were seen in PBMC in 3 and in ST in 2 patients of the 42 included in the trial. The 2 patients positive for coinfection in PBMC and who were undergoing combination antibiotic treatment were negative after six months, in contrast to the patient under placebo who remained positive after six months. A recent case report underlines the efficacy of chlamydial coinfection in ReA. A patient (with convincingly demonstrated coinfection-positive culture for
C.tr. and
C.pn. in SF, culture positive bronchoalveolar lavage (BAL) for
C.pn., real time polymerase chain reaction (RT-PCR) positive for both
Chlamydia spp. in SF, and RT-PCR positive in BAL for
C.pn.) achieved complete remission with the antibiotic combination of azithromycin plus rifampicin for three months, and another two months, after discontinuing of the medications for one month to induce the persistent organisms to return to their active developmental cycle [
6].
The results of the controlled trial and the case report are promising and support the causative role of Chlamydiae in arthritis, but those results also engender several questions: (1) Why, in the controlled trial, were about one-third of patients non-responders, and why was the rate of complete remission rather low? (2) Which is the most efficacious combination of antibiotics for treatment, and is there a need to optimize the dosing and duration of therapy? (3) Is it possible that in the controlled trial coinfections involving other bacteria associated with ReA were not identified, which may prevent the response to the treatment regimen?
The latter is an obvious possibility, given the most recent report of multiple coinfections of
Chlamydial spp.,
Mycoplasma, and
Ureaplasma in patients with post-venereal ReA [
30]. The case study of post-venereal ReA (
n = 22) assessed the presence of
C.tr.,
C.pn.,
M.hominis, and
U. urealyticum in samples of ST, SF, and PBMC at the time of synovectomy and after four-month antibiotic combination therapy (a combination of ciprofloxacin, tetracycline, and roxithromycin). Coinfections with two or three different bacteria were detected in 16/22 (72.7%) patients, most frequently in ST (8/17; 47.1%;
n = 3
C.tr. plus
C.pn.,
n = 4 C.tr. plus
M. hominis,
n = 1
C.tr. plus
C.pn. plus
M. hominis) and PBMC (10/22; 45.5%;
n = 6 C. tr. plus
U. urealyticum,
n = 1
C.tr. plus
C.pn.,
n = 1 C. tr. plus
M. hominis,
n = 2
C.tr. plus
C.pn. plus
M. hominis,
n = 1
C.tr. plus
C.pn. plus
U. urealyticum) samples [
30]. After synovectomy combined with antibiotic combination,
C.tr. was found in PBMC samples from 13/22 patients. At diagnosis, 7 patients were positive for
C.pn. and 6 for
M. hominis. After the therapy, 4 were still positive for
C.pn., and one patient remained positive for
M. hominis. Before therapy, 9 patients were positive for
U. urealyticum, all of whom became negative after therapy. The synovectomy probably contributed notably to the remission of patients because the hypertrophic ST containing infectious agents was removed [
30].
No intra-articular coinfection of
C.tr. and
N. gonorrhoeae has been reported to date, to the best of our knowledge, although urogenital coinfections are frequently described. However, other coinfections of
Chlamydiae and arthritogenic bacteria are of some relevance.
Borrelia burgdorferi is one of the most important arthritis-triggering organisms in western countries and is therefore important in the differential diagnosis of chlamydial arthritis. The simultaneous detection of DNA from
C.tr. and
B. burgdorferi in the SF of 6 patients with unexplained oligoarthritis was first described by Putschky and colleagues [
35]. Coincidental history of tick bite and
C.tr. positive in urogenital smears in 2/6 patients, positive serology for both bacteria in 1 patient, and positive
B. burgdorferi serology in combination with
C.tr. positive in urogenital smears in 1 patient, all support to some extent the suggestion that both bacteria may be causing the joint inflammation in these individual cases. Less convincing is the implicated role of coinfections involving
Yersinia enterocolitica,
C.pn. and
Mycoplasma pneumoniae in
Borrelia arthritis, which are based merely on serology and a lymphocyte transformation test [
36].
Thus, evidence for coinfections of Chlamydia spp. with one another and with Mycoplasma and/or Ureaplasma species in joints, based on molecular genetic testing, are available for ReA, uSpA, and undifferentiated oligoarthritis from case reports and from case series studies. The pathogenic and clinical implications will be discussed in general in analogy to evidence from other established bacterial coinfections and a few relevant in vitro studies.