Discussion
By studying 165,593 blood specimens from 13 USA hospitals, Tabak et al.
have shown that the median time to identify BSI pathogens using
traditional blood culture was 44.0 h, with a sensitivity of
approximately 70% in critically ill patients and even lower for
fastidious microorganisms (Tabak, et al., 2018). In this study, to
overcome the shortcomings of the blood culture in BSI diagnosis, we
developed a culture-independent ddPCR method to rapidly and accurately
identify A. baumannii and K. pneumonia in blood samples of
patients with suspected BSI. Our results showed that ddPCR could
identify A. baumannii and K. pneumonia in whole blood
samples within 4 h, with a specificity of 100% for each strain and
limit of detection at 0.93 copies/µL for A. baumannii and 0.27
copies/µL for K. pneumonia . Clinical validation in 170 patients
with suspected BSIs showed that ddPCR not only detected the positive
patients who were identified by blood culture but also detected the
patients who were missed by blood culture. Of note, compared with the
patients reported by both blood culture and ddPCR assay, the patients
reported by ddPCR alone had less severe clinical manifestations and
better clinical outcomes, suggesting that these patients benefit from
the guidance of ddPCR results in the early stage of BSI. Thus, ddPCR may
serve as a rapid and reliable method to identify causal pathogens in BSI
and to guide the treatment decisions in the early stage of BSI.
In BSI, the human immune system and antibiotic treatment kill invading
pathogens, leading to the release of the nucleic acids from the
pathogens into the blood, which become a part of circulating cell-free
DNA (cfDNA) (Lo, et al., 1999, Diehl, et al., 2008, Lu, et al., 2018,
Papadopoulos, 2020). Thus, the presence of specific pathogenic DNA in
cfDNA can reflect the presence of pathogens in the bloodstream.
Accumulating evidence have demonstrated the feasibility of
next-generation sequencing (NGS) of plasma cfDNA to identify the
pathogens in BSI (Blauwkamp, et al., 2019, Farnaes, et al., 2019,
Grumaz, et al., 2019, Rossoff, et al., 2019). However, the typical
turnaround time of 2 days and the high cost of NGS remain barriers for
cfDNA NGS in clinical practice (Long, et al., 2016, Simner, et al.,
2018, Chiu and Miller, 2019, Crawford, et al., 2019). In this study, we
took advantage of the ultra-high sensitivity of ddPCR and the
feasibility of cfDNA in pathogen identification to develop a ddPCR-based
method using cfDNA as the template. The turnaround time of ddPCR to
diagnosis was 4 h, which was significantly shorter than that of NGS
(2–3 d) or blood culture (90.6 ± 12.9 h in this study).
Wouters et al. have developed a ddPCR method to detect bacteria or fungi
using metagenomic DNA as the template and broad-range primer-probe sets;
however, the overall specificity in clinical validation was only 80%
(Wouters, et al., 2019). In this study, we used cfDNA as the template
and designed specific primer-probe sets for A. baumannii andK. pneumonia , respectively. We achieved 100% specificity in
well-characterized ATCC isolates of each strain, which was also higher
than other reported PCR-based methods, such as SeptiFast (50%,
(Warhurst, et al., 2015); 85.5%, (Korber, et al., 2017)), Magicplex
(29%) (Zboromyrska, et al., 2019), TAC assay (79.4%) (Zhang, et al.,
2018), T2Bacteria (90%) (Maki, 2019, Nguyen, et al., 2019), and cfDNA
NGS (93.7%) (Blauwkamp, et al., 2019).
The sensitivity of blood culture is typically lower than expected. The
blood culture positivity rate of septic patients was 51% over 22 years
in the United States (Martin, et al., 2003). Cheng et al. have reported
a 71.7% blood culture positivity rate in severe sepsis patients from
ten university hospitals in China (Cheng, et al., 2007). Similarly, 70%
of infected patients in a one-day international investigation in ICU
have been reported positive by blood culture test (Vincent, et al.,
2009). The low sensitivity of blood culture may attribute to the low
bacteria abundance in the blood, the antibiotic treatment before
sampling, and the culture techniques. Molecular detection methods are
less affected by these factors, thus usually have higher positivity
rates than blood culture. The positivity rates of different molecular
methods are 1.56 – 6.45-fold higher than that of blood culture (Long,
et al., 2016, Korber, et al., 2017, Farnaes, et al., 2019, Grumaz, et
al., 2019, Nguyen, et al., 2019). In the present study, the positivity
rates of A. baumannii and K. pneumonia were 6.8-fold
((23+4)/4) and 4.3-fold ((26+4)/7) higher than those of blood culture,
respectively. Thus, molecular detection methods may disclose the missed
diagnosis in blood culture, allowing timely diagnosis and appropriate
antibiotic treatment for the patients with BSI.
In this study, the 53 positive patients detected by ddPCR had typical
manifestations resulting from BSI, including body temperature greater
than 38.5℃, abnormally elevated serum levels of C-reactive protein and
procalcitonin, hemodynamic instability, and severe organ dysfunction. Of
note, the positive patients reported by ddPCR alone exhibited less
severe manifestations than those reported by both ddPCR and blood
culture, suggesting that ddPCR is more sensitive than blood culture for
early diagnosis of BSI.
In conclusion, we developed a novel ddPCR method to detect two major
pathogens in patients with suspected BSI. Clinical validation revealed
that our method outperformed the blood culture in specificity,
sensitivity, and turnaround time, serving as a promising method for
early and accurate diagnosis of BSI. However, in this pilot study, we
only included two major Gram-negative bacteria responsible for BSI. More
clinically important pathogens will be further investigated in future
studies.