Abstract
This prospective study evaluated the utility of the SeptiFast (SF) test in detecting 25 clinically important pathogens in 225 blood samples from 170 intensive care unit (ICU) patients with suspected sepsis after liver transplantation (LTX) or after other major abdominal surgery (non-LTX). SF yielded a significantly higher positivity rate in the LTX group (52.3%) than in the non-LTX group (30.5%; P = 0.0009). SF may be a powerful tool for the early diagnosis of bloodstream infections in LTX patients.
TEXT
Liver transplant recipients are at high risk of developing bloodstream infections (BSIs). Bacteremia has been documented in 20% to 40% of patients after liver transplantation (12, 13, 17), with associated mortality rates ranging from 15% to 36% (17). Invasive fungal infections such as candidiasis and invasive aspergillosis have also been reported in 5% to 42% of liver transplant recipients, with associated mortality rates of 25% to 71% (4, 10, 13). A commercially available multiplex PCR assay is the SeptiFast (SF) test (Roche Diagnostics, Mannheim, Germany), which is designed to detect the DNA of 25 clinically important bacteria and fungi in the blood within 6 h. In this study, the diagnostic performance of SF was for the first time compared to that of blood culture (BC) in patients with suspected sepsis after liver transplantation (LTX) or other major abdominal surgery (non-LTX).
This prospective observational study was carried out from May 2009 to April 2011 in the Department of General, Visceral, and Transplant Surgery of the University Hospital Essen. A total of 225 blood samples for blood culture analysis and SF testing were obtained in parallel from 170 intensive care unit (ICU) patients with suspected sepsis according to the criteria of the ACCP/SCCM (7). We separated patients into groups according to surgical procedure: those who had undergone liver transplantation and those who had undergone other major abdominal surgery. Patient characteristics are presented in Table 1.
Table 1.
Baseline demographic characteristics of 170 patients with suspected sepsis
| Characteristic | No. (%) of patients with:a |
|
|---|---|---|
| LTX group (n = 79 patients) | Non-LTX group (n = 91 patients) | |
| Mean age ± SD; range (yr) | 52.6 ± 10.9; 27–70 | 60.2 ± 13.2; 28–88 |
| Female | 28 (35.4) | 46 (50.5) |
| Cause of cirrhosis | ||
| Alcoholic | 21 (26.5) | |
| Infectious (hepatitis B/C) | 17 (21.5) | |
| NASH | 4 (5.0) | |
| Other (autoimmune, unknown) | 11 (13.9) | |
| Hepatocellular carcinoma | 12 (15.1) | |
| Primary sclerosing cholangitis | 7 (8.8) | |
| Acute liver failure | 4 (5.0) | |
| Liver cysts | 3 (3.7) | |
| Malignancy | 48 (60.7) | |
| Abdominal infection | 27 (29.6) | |
| Abdominal organ perforation | 12 (13.1) | |
| Colon ischemia | 4 (4.3) | |
Except for the first row (mean age), values are numbers (percentages) of patients. LTX, patients after liver transplantation; non-LTX, patients after other major abdominal surgery; NASH, nonalcoholic steatohepatitis.
Blood was collected by venipuncture or via a central venous catheter. At least 2 sets of BC bottles (Bactec 9240 Plus Aerobic/F and Anaerobic/F; Becton Dickinson, Heidelberg, Germany) were inoculated with 8 to 10 ml of blood per bottle and incubated for up to 5 days at 36°C.
SF is a European Council (CE)-marked in vitro diagnostic reagent kit for the detection of DNA from bacteria and fungi in human EDTA blood. For SF testing, two tubes of 3 ml EDTA blood each were drawn from the same venipuncture or catheter at the same time as the BCs were obtained. A detailed description of the SF workflow has been previously published (6). The spectrum of species that can be detected by SF is shown in Table 2. According to a publication by Richter et al., pathogens from skin flora were considered to be probable contaminants when detected in only one of two or more BCs (11). Categorical variables were compared with the chi-square test. Statistical analysis was performed with SAS version 9.2.
Table 2.
Spectrum of species that can be detected by SeptiFast
| Type of organism | Organism |
|---|---|
| Gram positive | Staphylococcus aureus |
| Coagulase-negative staphylococci | |
| Streptococcus pneumoniae | |
| Streptococcus spp. | |
| Enterococcus faecium | |
| Enterococcus faecalis | |
| Gram negative | Escherichia coli |
| Klebsiella (pneumoniae/oxytoca) | |
| Serratia marcescens | |
| Enterobacter (cloacae/aerogenes) | |
| Proteus mirabilis | |
| Pseudomonas aeruginosa | |
| Acinetobacter baumannii | |
| Stenotrophomonas maltophilia | |
| Fungus | Candida albicans |
| Candida tropicalis | |
| Candida parapsilosis | |
| Candida krusei | |
| Candida glabrata | |
| Aspergillus fumigatus |
The overall positivity rate for the entire cohort was 40.8% by SF and 35.5% by BC (P = 0.11). SF and BC yielded concordant negative results in 110 (48.9%) samples and concordant positive results in 57 (25.3%) (Table 3). The results from 35 (15.6%) paired blood samples were SF+/BC−, whereas the results from 23 (14.2%) samples were SF−/BC+. In 10 of these 23 SF−/BC+ samples, the microorganisms were considered to be contaminants. The positivity rate for SF was significantly higher in the LTX group (52.3%) than in the non-LTX group (30.5%; P = 0.0009). BC also yielded more positive results in the LTX group (43.9%) than in the non-LTX group (27.9%; P = 0.013). Sensitivity, specificity, and positive and negative predictive values for both groups are presented in Table 3. More than 50% of the SF−/BC+ samples were coagulase-negative staphylococci. Among those bacteria that SF detected in addition to those isolated by BC, Enterococcus spp. and Enterobacteriaceae were the most frequently identified. In 7 specimens (10.8%), Aspergillus fumigatus was detected only by SF (Table 4).
Table 3.
Diagnostic performance of blood culture and SeptiFast in 225 samples from patients with suspected sepsis
| Resulta | Valuec |
||
|---|---|---|---|
| Entire cohort (n = 225) | LTX (n = 107) | Non-LTX (n = 118) | |
| BC+ | 80 (35.5) | 47 (43.9) | 33 (27.9) |
| SF+ | 92 (40.8) | 56 (52.3) | 36 (30.5) |
| SF+/BC+ | 57 (25.3) | 38 (35.5) | 19 (16.1) |
| SF−/BC− | 110 (48.9) | 42 (39.3) | 68 (57.6) |
| SF+/BC− | 35 (15.6) | 18 (16.8) | 17 (14.4) |
| SF−/BC+ | 23 (10.2) | 9 (8.4) | 14 (11.8) |
| Sensitivityb of SF (%) | 71.2 | 80.9 | 57.6 |
| Specificityb of SF (%) | 75.9 | 70.0 | 80.0 |
| Positive predictive value of SF (%) | 62.0 | 67.9 | 52.8 |
| Negative predictive value of SF (%) | 82.7 | 82.3 | 82.9 |
SF, SeptiFast; BC, blood culture.
Considered BC as gold standard.
Values are the numbers (percentages) of samples unless otherwise noted. LTX, patients after liver transplantation; non-LTX, patients after major abdominal surgery.
Table 4.
Spectrum of pathogens detected by blood culture, SeptiFast, or both in 225 samples from patients with suspected sepsis
| Microorganism | No. (%) of isolatesa |
||
|---|---|---|---|
| SF+/BC+ | SF+/BC− | SF−/BC+ | |
| Gram-positive cocci | 21 (34.4) | 21 (32.3) | 16 (61.5) |
| Staphylococcus aureus | 5 (8.2) | 4 (6.2) | 1 (3.8) |
| Coagulase-negative staphylococci | 5 (8.2) | 3 (4.6) | 14 (53.8) |
| Enterococcus spp. | 11 (18.0) | 14 (21.5) | 1 (3.8) |
| Gram-negative rods | 34 (55.7) | 30 (46.2) | 5 (19.2) |
| Enterobacteriaceae | 25 (41.0) | 22 (33.8) | 4 (15.4) |
| Pseudomonas aeruginosa | 8 (13.1) | 6 (9.2) | 1 (3.8) |
| Stenotrophomonas maltophilia | 1 (1.6) | 2 (3.1) | 0 (0.0) |
| Fungi | 5 (9.8) | 11 (16.9) | 3 (11.5) |
| Candida spp. | 6 (9.8) | 4 (6.2) | 3 (11.5) |
| Aspergillus fumigatus | 0 (0.0) | 7 (10.8) | 0 (0.0) |
| Otherb | 0 (0.0) | 3 (4.6) | 2 (7.7) |
| Totalc | 61 (100) | 65 (100) | 26 (100) |
SF, SeptiFast; BC, blood culture.
The other pathogens detected in SF+/BC− samples included Acinetobacter baumannii (detected once) and Streptococcus spp. (detected twice), and the other pathogens detected in SF−/BC+ samples included Aeromonas hydrophila and Cryptococcus neoformans.
Number of samples in which more than one organism was detected: SF+/BC+, 4; SF+/BC−, 30; SF−/BC+, 3.
The performance of SF has been evaluated in a variety of patient populations, but information about its performance in transplant patients is scarce. Our results are in agreement with previous reports which also showed that the rates of positive findings are higher for SF than for BC in patients with febrile neutropenia, in newborns and children, and in septic ICU patients (5, 8, 9, 16, 18, 19).
The rapid availability of the results of SF may influence early therapeutic decisions for septic patients, particularly when potentially more-resistant pathogens are detected. The presence of these pathogens, which are not always covered by initial, empirical antibiotic therapy, is an important cause of treatment failure in septic patients. Another reason for using SF to analyze blood samples from critically ill patients such as LTX patients is that the detection of yeast and filamentous fungi in blood is limited to their specific growth abilities. Invasive aspergillosis (IA) occurs in 1% to 10% of LTX patients (2, 14), with a high mortality rate (15). SF detected A. fumigatus in blood from 4 LTX and 3 non-LTX patients. According to the revised EORTC/MSG criteria (3), 5 of these 7 patients had probable IA and 2 had proven IA. In neutropenic patients, two studies highlighted the value of SF in the rapid diagnosis of IA. When data from these studies are taken together, SF detected DNA from A. fumigatus in all 9 patients with probable IA, whereas the results of BC were negative (9, 16).
In contrast to all previous performed studies, the present study found that the overall rate of positive results by SF was very high (52.3%) in the LTX group, even though a high percentage of the patients included in the study had received empirical or preemptive antimicrobial therapy. This might lead to the hypothesis that in septic LTX patients, the bacterial load is higher than in the non-LTX group. Therefore, LTX patients with suspected sepsis seem to be a good cohort for the use of SF for rapid detection of pathogens in the bloodstream.
One limitation of the present study is that we did not attempt to correlate the SF+/BC− results with anti-infective treatment, the response to a specific therapy, or clinical outcome. However, findings of a multicenter trial comparing BC with SF in severe human sepsis showed that pathogen detection by SF was positively correlated with biomarkers of host response and disease severity even if the results of BC were negative (1).
SF was more sensitive than BC in LTX patients, a finding suggesting that LTX patients with suspected sepsis are a good cohort for the rapid and improved diagnosis of bloodstream infections by SF.
ACKNOWLEDGMENTS
We thank the staff members of the Institute of Medical Microbiology, especially those of the Laboratory of Molecular Microbiology, for excellent technical assistance.
P.-M. Rath and J. Steinmann have served on speakers' bureaus for Roche Diagnostics Germany. F. Saner has served on speakers' bureaus for Pfizer, MSD, Fresenius, and Gilead.
Footnotes
Published ahead of print 4 April 2012
REFERENCES
- 1. Bloos F, et al. 2010. A multicenter trial to compare blood culture with polymerase chain reaction in severe human sepsis. Intensive Care Med. 36:241–247 [DOI] [PubMed] [Google Scholar]
- 2. Brown RS, Jr, et al. 1996. Incidence and significance of Aspergillus cultures following liver and kidney transplantation. Transplantation 61:666–669 [DOI] [PubMed] [Google Scholar]
- 3. De Pauw B, et al. 2008. Revised definitions of invasive fungal disease from the European Organization for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) Consensus Group. Clin. Infect. Dis. 46:1813–1821 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Kusne S, et al. 1988. Fungal infections after liver transplantation. Transplant. Proc. 20:650–651 [PMC free article] [PubMed] [Google Scholar]
- 5. Lamoth F, et al. 2010. Multiplex blood PCR in combination with blood cultures for improvement of microbiological documentation of infection in febrile neutropenia. J. Clin. Microbiol. 48:3510–3516 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Lehmann LE, et al. 2008. A multiplex real-time PCR assay for rapid detection and differentiation of 25 bacterial and fungal pathogens from whole blood samples. Med. Microbiol. Immunol. 197:313–324 [DOI] [PubMed] [Google Scholar]
- 7. Levy MM, et al. 2003. 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit. Care Med. 31:1250–1256 [DOI] [PubMed] [Google Scholar]
- 8. Lucignano B, et al. 2011. Multiplex PCR allows rapid and accurate diagnosis of bloodstream infections in newborns and children with suspected sepsis. J. Clin. Microbiol. 49:2252–2258 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Mancini N, et al. 2008. Molecular diagnosis of sepsis in neutropenic patients with haematological malignancies. J. Med. Microbiol. 57:601–604 [DOI] [PubMed] [Google Scholar]
- 10. Neofytos D, et al. 2010. Epidemiology and outcome of invasive fungal infections in solid organ transplant recipients. Transpl. Infect. Dis. 12:220–229 [DOI] [PubMed] [Google Scholar]
- 11. Richter S, et al. 2002. Minimizing the workup of blood culture contaminants: implementation and evaluation of a laboratory-based algorithm. J. Clin. Microbiol. 40:2437–2444 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Saner FH, et al. 2008. Pulmonary and blood stream infections in adult living donor and cadaveric liver transplant patients. Transplantation 85:1564–1568 [DOI] [PubMed] [Google Scholar]
- 13. Singh N, Gayowski T, Wagener M, Yu VL. 1994. Infectious complications in liver transplant recipients on tacrolimus. Prospective analysis of 88 consecutive liver transplants. Transplantation 58:774–778 [PubMed] [Google Scholar]
- 14. Singh N, Gayowski T, Wagener MM, Doyle H, Marino IR. 1997. Invasive fungal infections in liver transplant recipients receiving tacrolimus as the primary immunosuppressive agent. Clin. Infect. Dis. 24:179–184 [DOI] [PubMed] [Google Scholar]
- 15. Singh N, et al. 2006. Combination of voriconazole and caspofungin as primary therapy for invasive aspergillosis in solid organ transplant recipients: a prospective, multicenter, observational study. Transplantation 81:320–326 [DOI] [PubMed] [Google Scholar]
- 16. von Lilienfeld-Toal M, et al. 2009. Utility of a commercially available multiplex real-time PCR assay to detect bacterial and fungal pathogens in febrile neutropenia. J. Clin. Microbiol. 47:2405–2410 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Wade JJ, et al. 1995. Bacterial and fungal infections after liver transplantation: an analysis of 284 patients. Hepatology 21:1328–1336 [DOI] [PubMed] [Google Scholar]
- 18. Westh H, et al. 2009. Multiplex real-time PCR and blood culture for identification of bloodstream pathogens in patients with suspected sepsis. Clin. Microbiol. Infect. 15:544–551 [DOI] [PubMed] [Google Scholar]
- 19. Yanagihara K, et al. 2010. Evaluation of pathogen detection from clinical samples by real-time polymerase chain reaction using a sepsis pathogen DNA detection kit. Crit. Care 14:R159. [DOI] [PMC free article] [PubMed] [Google Scholar]