Culture and Real-time Polymerase Chain reaction sensitivity in the diagnosis of invasive meningococcal disease: Does culture miss less severe cases?

Background: Invasive meningococcal disease (IMD) is a highly lethal disease. Diagnosis is commonly performed by culture or Realtime-PCR (qPCR).
Aims: Our aim was to evaluate, retrospectively, whether culture positivity correlates with higher bacterial load and fatal outcome. Our secondary aim was to compare culture and qPCR sensitivity.
Methods: The National Register for Molecular Surveillance was used as data source. Cycle threshold (CT), known to be inversely correlated with bacterial load, was used to compare bacterial load in different samples.
Results: Three-hundred-thirteen patients were found positive for Neisseria meningitidis by qPCR, or culture, or both; 41 died (case fatality rate 13.1%); 128/143 (89.5%) blood samples and 138/144 (95.8%) CSF were positive by qPCR, 37/143 (25.9%) blood samples and 45/144 (31.2%) CSF were also positive in culture. qPCR was 3.5 times (blood) or 3.1 times (CSF) more sensitive than culture in achieving a laboratory diagnosis of IMD (OR 24.4; 95% CI 12.2-49.8; p < .10-4; Cohen's κ 0.08 for blood and OR 49.0; 95% CI 19.1-133.4; p<10-4; Cohen's κ 0.02; for CSF). Positivity of culture did not correlate with higher bacterial loads in blood (mean CT 27.7±5.71, and CT 28.1±6.03, p = 0.739 respectively in culture positive or negative samples) or in CSF (mean CT 23.1±4.9 and 24.7±5.4 respectively in positive or negative CSF samples, p = 0.11).CT values in blood from patients who died were significantly lower than in patients who survived (respectively mean 18.0, range 14-23 and mean 29.6, range 16-39; p<10-17). No deaths occurred in patients with CT in blood over 23. Positive blood cultures were found in 10/25 (40%) patients who died and in 32/163 (19.6%) patients who survived, p = 0.036, OR 2.73; 95% CL 1.025-7.215), however 60% of deaths would have remained undiagnosed with the use of culture only.
Conclusions: In conclusion our study demonstrated that qPCR is significantly (at least 3 times) more sensitive than culture in the laboratory confirmation of IMD. The study also demonstrated that culture negativity is not associated with lower bacterial loads and with less severe cases. On the other side, in patients with sepsis, qPCR can predict fatal outcome since higher bacterial load, evaluated by qPCR, appears strictly associated with most severe cases and fatal outcome. The study also showed that molecular techniques such as qPCR can provide a valuable addition to the proportion of diagnosed and serotyped cases of IMD.
Competing Interests: SG received a scholarship from GSK for meningococcal epidemiology; GSK did not participate in study design, data analysis, manuscript writing and in any other aspects related to the present work. This does not alter our adherence to PLOS ONE policies on sharing data and materials. The authors have no other conflict of interest.