Realtime PCR is more sensitive than multiplex PCR for diagnosis and serotyping in children with culture negative pneumococcal invasive disease

Abstract

Background: Pneumococcal serotyping is usually performed by Quellung reaction, considered the gold standard test. However the method cannot be used on culture-negative samples. Molecular methods can be a useful alternative. The aim of the study was to evaluate the use of Multiplex-sequential-PCR (MS-PCR) or Realtime-PCR on blood samples for diagnosis and serotyping of invasive pneumococcal disease (IPD) in a pediatric clinical setting.

Methodology/principal findings: Sensitivity and specificity of MS-PCR and Realtime-PCR have been evaluated both on 46 well characterized pneumococcal isolates and on 67 clinical samples from children with culture-negative IPD. No difference in sensitivity and specificity between MS-PCR and Realtime PCR was found when the methods were used on isolates: both methods could type 100% isolates and the results were always consistent with culture-based methods. On the contrary, when used on clinical samples 43/67 (64.2%) were typeable by MS-PCR and 61/67 (91.0%) by Realtime-PCR (p = 0.0004,K Cohen 0.3, McNemar's p<0.001). Non-typeability by MS-PCR was associated in 18/20 cases (90.0%) with low bacterial load. The difference between the two methods was present both when they were used on normally sterile fluids (respectively 31/33 (93.9%) typeable samples for Realtime-PCR and 24/33 (72.7%) for MS-PCR, p = 0.047, 95%CL 0.03-0.98; K Cohen 0.3; McNemar's p = 0.0016) and when they were used on nasopharyngeal swabs (respectively 30/34 (88.2%) typeable samples for Realtime-PCR and 19/34 (55.9%) for MS-PCR, p = 0.007, 95%CL 0.04-0.66); the presence of multiple pneumococcal serotypes in nasopharyngeal swabs was found more frequently by Realtime PCR (19/30; 63.3%) than by Multiplex-sequential PCR (3/19; 15.8%; p = 0.003;95%CL 1.87-39.97).

Conclusions/significance: In conclusion, both MS-PCR and Realtime PCR can be used for pneumococcal serotyping of most serotypes/serogroups directly on clinical samples from culture-negative patients but Realtime-PCR appears more sensitive.

Figure 1. Number of typeable or non-typeable samples obtained from patients with culture-negative invasive pneumococcal infections as evidenced by Multiplex Sequential PCR or Realtime-PCR (sensitivity of Realtime vs Multiplex Sequential PCR p = 0.0004; 95%CL 1.98–17.05).

Figure 2. Presence of single or multiple pneumococcal serogroups/serotypes in nasopharyngeal swabs as evidenced by Multiplex Sequential PCR or Realtime-PCR (frequency of multiple serotypes/serogroup evaluated with Realtime PCR vs Multiple Sequential PCR p = 0.003; 95%CL = 1.87–39.97).