Background
Chlamydia psittaci is an obligatory intra-cellular Gram-negative bacterium that typically infects birds, but could occasionally cause psittacosis in humans when contaminated aerosols from infected birds are inhaled.
Chlamydia psittaci pneumonia in humans is underestimated due to low awareness of the disease and atypical clinical presentation in majority of the cases [
1‐
3]. The low sensitivity and complex procedure of
Chlamydia psittaci culture causes it hardly routinely performed in most diagnostic laboratories. Other laboratory testing included serological assay and polymerase chain reaction (PCR) based methods, but both have questionable sensitivity and specificity [
2]. Untargeted metagenomic next-generation sequencing (mNGS) has been increasingly used in the diagnosis of infectious diseases, particularly when conventional diagnostic approaches have limitations [
4]. Here we report 5 cases of
Chlamydia psittaci pneumonia, in which the diagnosis was established with mNGS. All together, mNGS was conducted in a total of 120 pneumonia cases in the index period. Demographical and basic clinical features of the 5 cases are summarized in Table
1.
Table 1Demographical and basic clinical features of the 5 patients
1 | Female | 81 | BALF | hypertension, diabetes, coronary artery diseases | WBC 7.71*10^9, NE% 69.8, NE 5.38*10^9, LN% 17.5↓, LN 1.35*10^9 | 53 | <0.05 | 74 | Candida albicans (182) Staphylococcus capitis (12) Chlamydia psittaci (6) Human betaherpesvirus 5 (2) |
2 | Male | 45 | BALF | diabetes | WBC 8.22*10^9, NE% 93.1↑, NE 7.65*10^9↑, LN% 4.0↓, LN 0.33*10^9↓ | 226 | 1.712 | 42 | Chlamydia psittaci (225) Lautropia mirabilis (28) Leuconostoc lactis (19) Rothia mucilaginosa (19) Rothia dentocariosa (12) Streptococcus parasanguinis (10) Actinomyces odontolyticus (8) Streptococcus mitis (4) |
3 | Female | 85 | Lung tissue | none | WBC 6.38*10^9, NE% 88.9, NE 5.67*10^9, LN% 8.0↓, LN 0.51*10^9 | 184 | 1.64 | 64 | Chlamydia psittaci (48) Staphylococcus epidermidis (1) |
4 | Female | 66 | Lung tissue | none | WBC 5.47*10^9, NE% 69.8, NE 3.82*10^9, LN% 19.4↓, LN 1.06*10^9↓ | 124 | 0.13 | 17 | Chlamydia psittaci (205) Corynebacterium striatum (1) Klebsiella pneumoniae (1) |
5 | Female | 61 | Lung tissue | none | WBC 6.88*10^9, NE% 68.4, NE 4.70*10^9, LN% 22.7↓, LN 1.56*10^9↓ | 96.7 | <0.05 | 37 | Chlamydia psittaci (2) |
Discussion and conclusions
Chlamydia psittaci can be classified into 10 genotypes, with varying preference for host species [
5]. Genotype A and E could infect humans. After entry via contaminated aerosols,
Chlamydia psittaci spreads to the reticuloendothelial system. The lungs are the most common sites of
Chlamydia psittaci infection.
Chlamydia psittaci pneumonia is estimated to account for approximately 1% of community-acquired pneumonia cases [
3,
6]. Symptoms mimics that of influenza, and typically include fever, fatigue, headache, myalgia, and coughing [
7,
8].
In 4 out of the 5 cases in this report, patients had contacts with birds (parrot and pigeon) or poultry, suggesting the need to investigate such exposure upon suspected cases. Two out of the 5 patients had diabetes; and the other 3 were otherwise healthy, suggesting that Chlamydia psittaci could infect human subjects regardless of underlying diseases.
Chlamydia psittaci infection tends to be overlooked due to relatively low awareness by physicians. Laboratory testing for
Chlamydia psittaci includes culture, serological assay, and PCR. Culture is time consuming, and most formidably, requires P3 facility [
6]. Serological tests are only appropriate for retrospective diagnosis because sera from the both acute and convalescent phase of the illness are required [
9]. PCR-based testing is the most specific and fastest method but only sensitive in the acute phases of the infection [
2]. The current case series indicated that mNGS could be used to diagnose
Chlamydia psittaci infection. If using a set of universally accepted standards, mNGS could even provide semiquantitative information (based on sequence reads) about the load of
Chlamydia psittaci, such information could be critically important in determining whether a specific microbe is the causative pathogen(s) in polymicrobial samples. The limitations of mNGS analysis, for infection with a
Chlamydia psittaci or any other agent, include host background sequences. In future studies, targeted sequencing and host depletion methods could be used to minimize the human host background, workflow quality control procedures could be optimized to reduce false positives. It is also likely that
Chlamydia psittaci is merely present in the sample and not the culprit of infection. The fact that many other pathogens, including
Candida albicans, have been identified in the 5 cases illustrate the complexity and a need to integrate the mNGS results into the overall clinical scenario. Lack of verification with serologic tests and/or culture is a significant limitation in the current study. These findings therefore must be interpreted with caution.
In the current series, mNGS was conducted only after initial empirical antibiotic treatment failed to control the infection in 4 out of the 5 cases. Considering the feasibility and cost of mNGS, we believed the timing of mNGS is appropriate, and recommend mNGS testing only if patients do not respond to treatments against other more common causes.
The specific reads for Chlamydia psittaci ranged from 2 to 225 in the 5 cases. Comparison across cases is not possible since the assay was conducted by several different companies. In case #1 and #5, the detected Chlamydia psittaci reads were 6 and 2, respectively. We speculate that the relatively low specific reads of Chlamydia psittaci in Case #1 may reflect the therapeutic effects of moxifloxacin, albeit not adequate. In Case #5, the low specific reads of Chlamydia psittaci may reflect the loss of biological activity of pathogens and degradation of nucleic acid during the process of sample collection and transportation or the treatments that the patient received prior to sample collection.
Recommended treatment for
Chlamydia psittaci pneumonia included tetracycline, macrolide and quinolones [
10]. Treatment must continue for at least 10–14 days to prevent relapse. In case #1, moxifloxacin was used as an initial treatment for 3 days, but no significant improvement was obtained. In our opinion, the lack of response to moxifloxacin in this case could be due to several reasons, including: 1) possible superinfection by other agents that could be readily controlled by doxycycline; 2) relative insensitivity of the
Chlamydia psittaci isolate in this specific case to moxifloxacin. Indeed, recommended first-line treatment of
Chlamydia psittaci pneumonia is doxycycline and not quinolones [
11].
In conclusion, the current series suggested that mNGS could be used in diagnosing Chlamydia psittaci infection. This preliminary finding should be examined with diagnostic trial in the future.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.