Legionella Pneumophila Infections among Egyptian Patients: Updated Diagnostic Tests and Evaluation

• Urinary antigen test for L. pneumophila and q RT PCR testing are promising method that can be used for rapid identification of Legionella infections.
• Improvement in the diagnosis of legionella will help to better identification of the prevalence of Legionnaire`s disease.
• The development of standardized PCR assays will be major advances in Legionella diagnostics. But these techniques must be interpreted cautiously and hand-in-hand with culture .

Table 1: Relation between q RT PCR and respiratory sample culture

In this study, the prevalence of Legionnaires’ disease (LD) in Zagazig University Hospitals by using the urinary antigen testing, the molecular technique (q RT PCR mip gene) and the culture of the lower respiratory specimens was 17%, 15 % and 12% respectively. It is reported that Legionella Pneumophila comprises 2-9% of the cases of community‐acquired pneumonia (CAP) [9-13]. The mean incidence of Legionella infection in CAP globally was as follow; Africa 1.6% , Eastern Mediterranean 9.7%, Europe 4.5%, South-East Asian 4.3%, USA 1.5% and Western Pacific 4.5% [14-16]. The large heterogeneity in the incidence estimates for Legionella infections is attributed to many factors including variable definitions of CAP, patients’ characteristics, diagnostic methods, and criteria used for diagnosis [15].

The current gold standard test used for the diagnosis of LD is the lower respiratory specimen culture although its use remains challenging, as the culture systems have some limitations, including delay of the results, sample non-accessibility, and low sensitivity, especially after initiation of antibiotics [17]. To hasten the diagnosis and improve the sensitivity, two diagnostic tools were evaluated, q RT PCR mip gene and urinary antigen test (UAT).

In the present study, out of 100 samples taken from patients with confirmed pneumonia, Legionella pneumophilia was detected in 12% of the lower respiratory specimen cultures which was in the positivity range of other studies; Ricci et al. found that culture positivity rate was 9.9% [18]. Moreover, Cloud et al. found a culture positivity rate of 14.6% [19]. On the other hand, several studies showed lower positivity rates 2.9% by Kese et al. [2], 5% by Nageeb et al. [20] and 2.8% by Peci et al. [21]. These differences in positivity of lower respiratory specimen culture results would be attributed to differences in the disease endemicity in different localities and exposure to antimicrobials.

The current study showed that Legionella was detected in 15% of the total samples by molecular technique (q RT PCR mip gene). This improvement in Legionella detection by using PCR goes in accordance with different studies. Recci et al. reported that PCR was positive in 53 samples (15%) out of 354 samples, while respiratory specimen culture was positive only in 35 samples (10%) [18]. While Kese et al. reported that out of 3038 samples, PCR detected 128 (4.2%) positive samples, while respiratory specimen culture detected only 88 (3%) positive samples [2].

It was noted that, out of the 15 positive samples by the PCR, five samples were tested negative for the respiratory samples’ cultures. This could be interpreted by the detection of non-viable dead or non-cultivable legionella by the PCR [22], as well as contamination of the PCR reagents with the microbial DNA, which will result in false-positive results [23]. Additionally, two of the 12 culture-positive samples, the PCR results were negative. This may be due to inhibition of the PCR reaction, sequence variability underlying the primers and probes, or the presence of Legionella species in quantities less than the limit of detection of the assay [24]. Therefore, a negative PCR result indicates the absence of detectable Legionella`s DNA in the specimen but does not rule-out the infection.  The presence of two cases, which were diagnosed only by the culture and missed by the molecular technique and also 5 cases with positive PCR and negative culture results makes the interpretation of the PCR should be done cautiously and hand-in-hand with culture which must associates this technique. For evaluating the PCR assay compared to culture, the results showed a sensitivity of 83.3%, a specificity of 94.3%, a positive predictive value of 66.6%, a negative predictive value of 97.6% and an accuracy of 93.0% (Table 3). Moreover, the results obtained by the molecular technique correlated well with those obtained by culture (Kappa co-efficient 0.70±0.008 p<0.001). (Table 3). Good concordance with these results was found in other studies using conventional or automated PCR techniques. Ricci et al., showed that, overall concordance between the culture and the PCR was significant (k = 0.75; p < 0.0001). However, the sensitivity was 63.6%, which was lower than our study, while the specificity was 100%, along with the PPV and the NPV were 100% and 93.7%, respectively [18]. Mérault et al. showed that, the PCR and culture results were concordant for 182 out of 209 tested samples. The kappa coefficient was 0.57, which might be considered good agreement. The specificity was 84.1%, the sensitivity was 84.6%, the positive predictive value (PPV) was 55% and the negative predictive value was 95.5% [25].  The PCR had a 100% sensitivity reported by two studies with specificities of 93% and 100% [19-26]. Additionally, a retrospective study conducted in the Netherlands reported a 92% sensitivity and a 98% specificity for a PCR targeting the mip gene [27].

In the current study, the UAT for the legionella was positive in 17 samples out of the 100 tested samples for pneumonic patients (17%). While by using respiratory specimen culture technique only 12 samples (12%) were found to be positive. Similar results were shown by Ricci et al., which recorded 40 positive samples for Legionella by UAT technique out of 278 pneumonic patients (14.3%) and 21samples of them were also positive by Sputum culture while the total positive cases detected by the culture were 35 out of 354 cases (9.9%) [18]. It is apparent from our study that, the positive rate of the UAT (17%) exceeds that for the culture (12%); this was also found by Chen et al. as the positivity rate in pneumonic patients by culture was (0.4%) while by UAT positivity rate for Legionella was (2.7%) [28].

For evaluation of UAT assay, it has a sensitivity of 100%, a specificity of 94.3%, a positive predictive value of 70.6%, a negative predictive value of 100% and an accuracy of 95%. In addition, it has a significant correlation with the culture (Kappa agreement 0.79 ± 0.009, p value < 0.001). (Table3). Similar results were obtained by another study [29]. On the other hand, 5 out of 100 patients (5%) had negative cultures but positive UAT. This might be explained by the fact that, the urinary secretion of the Legionella antigen occurs within 2-3 days after the infection and lasting for months or even year after the end of infection [30]. Another explanation is the occurrence of non-specific signals, possibly due to the presence in urine of immune complexes that interact with the test and give false positive results [31,32]. For this reason, a single testing to assess efficacy is limited in most clinical practice. Combining other methods for simultaneous testing may maximize its accuracy and further evaluation for predicting the effect of UAT results in clinical practice is needed [33].

In the present study both the sensitivity and the negative predictive values of the UAT were 100%.  This is due to the usage of Biotest Legionella urine antigen Enzyme Immunoassay (EIA) kit that can detect the antigens of all L. pneumophila serogroups [34]. Enhanced sensitivity was also achieved by using novel UAT for legionella kit (LAC-116), that can detect serogroups of legionella other than the subgroup 1 [35]. This was in agreement with our results. 100 % sensitivity was also recorded by other study which evaluated UAT for legionella [36]. On the contrast of the previous results several studies recorded sensitivity and negative predictive values less than 80% [21, 37]. Ahmed et al. stated that there was a limitation of UAT as a diagnostic tool (sensitivity of 70-80%) and it could miss the non-serogroup-1 cases. Additionally, it concluded that PCR seems to have a higher accuracy among non-invasive testing, but no method alone is accurate enough to safely rule-in or out LD diagnosis [38]. This difference in the sensitivity may be attributed to the Legionella urinary antigen kits used in these studies which were specific for L. pneumophila serogroup-1 antigen only and so there was failure to detect the disease caused by the other serogroup. The L.pneumophila serogroup-1 causes 50-80% of Legionnaires’ disease; therefore, 20-50% of the case remain undiagnosed [38, 39].

Timely rapid identification of the causative pathogens is important for adequate antimicrobial therapy especially for pneumonic patients. The studied PCR and UAT techniques showed shorter turnaround time than that for respiratory specimen culture which considered a strong additive for management of these cases.

Conclusion

In conclusion urinary antigen test for L. pneumophila and q RT PCR testing are promising method that can be used for rapid identification of Legionella infections in pneumonic patients as both techniques correlated well with the gold standard technique, offering the possibility of identifying more positive cases that may be missed by lower respiratory specimen culture method and to rule out infection. Improvement in the diagnosis of legionella will help to better identification of the prevalence of Legionnaire`s disease. This improvement depends on the use of urinary antigen tests capable of detection of different serotypes of legionella; in addition, the development of standardized PCR assays will be major advances in Legionella diagnostics. But these techniques must be interpreted cautiously and hand-in-hand with culture .

Conflict of interest:None.

Funding: None.

:Ethical considerations

This study was approved by the Institutional Research Board, faculty of medicine, Zagazig University (approval No. 10374). This was done after consent the patients for their sharing in the study.