Adjuvant role of SeptiFast to improve the diagnosis of sepsis in a large cohort of hematological patients


Febrile neutropenia and sepsis are common and life-threatening complications in hematological diseases. This study was performed retrospectively in 514 patients treated for febrile neutropenia at our institute, to investigate the clinical usefulness of a molecular tool, LightCycler® SeptiFast test (SF), to promptly recognize pathogens causing sepsis in hematological patients. We collected 1837 blood samples of 514 consecutive hematological patients. The time of processing is short. Overall, 757 microorganisms in 663 episodes were detected by molecular test and standard blood cultures (BC): 73.6% Gram-positive bacteria, 23.9% Gram-negative bacteria, and 2.5% fungal species. This large analysis demonstrated a significant episode-to episode agreement (71.9%) between the two methods, higher in negative samples (89.14%), and a specificity of 75.89%. Clinical variables that gave a statistically significant contribution to their concordance were absolute neutrophil count, ongoing antimicrobial therapy, timing of test execution, and organ localization of infection. The large analysis highlights the potential of molecular-based assays directly performed on blood samples, especially if implementing the detection of antibiotic resistance genes, which was lacking in the used study.


Infections in patients with hematological malignancies are frequent and life-threatening complications [1]. Severe sepsis [2] and multiple organ failure are frequent cause of morbidity and mortality after hematopoietic stem cell transplantation (HSCT). Morbidity and mortality rates are even higher when Gram negative multidrug-resistant (GN-MDR) pathogens are involved. Recent improvement in supportive care have resulted in a considerable reduction of the treatment-related mortality [3], even though more effective prophylactic strategies and interventions are warranted [2, 4, 5]. Blood cultures (BC) are generally positive in only one third of cases; the culturing process and pathogen identification are lengthy, delaying the start of a pathogen targeted therapy. Thereby, a fast tool to rapidly detect pathogens causing sepsis could improve the diagnostic algorithm for sepsis [6, 7].

A preliminary experience with LightCycler® SeptiFast was previously conducted at our institution, demonstrating a good agreement with standard BC (83%) [8]. Others have confirmed its capacity to quickly detect bloodstream infections in immunosuppressed patients [9,10,11,12] and its usefulness in the context of sepsis [7, 13] or febrile neutropenia [14].

We tested the ability of LightCycler® SeptiFast to hasten the diagnostic process of sepsis, first by comparing it with standard blood culture considered as gold standard, and then evaluating its potential as a more sensitive molecular method.

Materials and methods

This study was performed retrospectively by using laboratory and clinical data of hematological patients treated at the Hematology and BMT Unit of our institution, from 2009 to 2013. Antimicrobial prophylaxis and treatment of infectious complications were administered according to the institutional guidelines [15,16,17], following international recommendations upon written informed consent for all procedures.

Following the internal guidelines, samples were taken at fever onset and every 3 days at fever persistence. A new fever onset after a 72-h of apyrexia was considered as a new febrile episode. Clinical events, laboratory, and imaging results, were analysed retrospectively. At least two sets of aerobic and anaerobic bottles were taken during the acute febrile episode (elevation of body temperature >38 °C) from a central venous catheter (CVC) or a peripheral vein. Additionally, another sample of blood was concomitantly taken in a EDTA tube for the molecular test, performed in real time on fresh samples.

Cultural approach

Both aerobic and anaerobic samples were processed in a BacT/Alert 3D automated BC system. The monitoring of CO2 production within each bottle was performed every 10 min, 24 h per day. All positive bottles were removed from the instrument, and an aliquot was used for Gram stain and culture on solid media for analysis. Antibiotic susceptibility was determined with the VITEK 2 system (bioMérieux), requiring not less than 24–36 h.

LightCycler® SeptiFast test

The molecular assay required 1.5 ml K-EDTA-treated uncultured blood for each sample. SeptiFast (Roche molecular systems) is capable to detect 25 among the most common species isolated in BC (Table 1). The test makes use of dual fluorescent resonance energy transfer (FRET) probes directed against the species-specific internal transcribed spacer (ITS) regions, which represent non-coding sequences interspaced among highly conserved ribosomal RNA of bacteria and fungi. Results are obtained within two working days with a median time of 6 h. A more detailed description of this molecular test was reported in our previous experience [8].

Table 1 LightCycler® SeptiFast test (Roche molecular systems): microbial panel

Statistical analysis

Data are described as raw number (percentage) when reporting frequencies and mean ± standard deviation when reporting normally distributed continuous variables. We calculated the specificity, sensitivity, positive (PPV), and negative (NPV) predictive value of the test, referencing it against the gold standard (blood cultures) by using a 2 × 2 crosstab. If a polymicrobial finding was detected by one or both molecular and culture tests, the above mentioned analysis was calculated on the basis of the concordance of at least one of the detected pathogens [8]. We reviewed infections within 30 days of death to examine the proximate relationship of any infection to death (infection-related mortality, IRM) [18]. We analysed the possible influence on the concordance between tests of nine variables (underlying disease, neutropenia grade, ongoing antimicrobial therapy, test timing, and site of infection) using a logistic regression model. Results of the logistic regression are described as odds ratio (OR), p value, 95% confidence intervals (CI). We considered a significant p-value lower than 0.05. The statistical analyses of the study were conducted with IBM SPSS for Windows (IBM Corp. Version 20.0, Armonk, NY).


A total of 1837 episodes from 514 patients, affected by febrile neutropenia, were collected between 2009 and 2013 (Table 2). The majority of patients had a severe neutropenia (60.4%) in the setting of an acute leukemia (60.5%).

Table 2 Patient baseline characteristics

Only 548 of the samples were collected at the onset of fever, while 960 were taken during an ongoing fever. The majority of tests (1200) were executed during the course of antimicrobial therapy. Only a small number of patients were tested in an afebrile state, mostly because of internal guidelines regarding CVC surveillance at hospital admission.

Together, SeptiFast and standard BC identified a total of 757 microorganisms in 663 out of the total 1837 episodes: 73.6% were Gram-positive bacterial species, 23.9% were Gram-negative bacterial species, and 2.5% were fungal species. SeptiFast gave positive results in 28.3% of cases, whereas 17.26% of total samples became culture positive (Fig. 1).

Fig. 1

Concordance between SeptiFast and blood cultures. Discordant positivity to SeptiFast/blood culture includes cases of isolation of different germs by the two methods. Concordance rate in microbial identification: 71.9% (discordant positivity not included). SF SeptiFast, BC blood culture

Out of the positive molecular samples, 464 were monomicrobial and 56 were polymicrobial. Among the culture-positive events, only 18 were polymicrobial. Overall, both methods detected a total of 580 and 333 microorganisms respectively (Fig. 2).

Fig. 2

Microorganism identification. The figure shows the absolute number of fungi, Gram-negative bacteria, and Gram-positive bacteria isolated by blood culture and SeptiFast. The striped columns does not respect the axis label. SF SeptiFast, BC blood culture

When compared to blood cultures, we found SeptiFast to have a sensitivity of 50.69%, a specificity of 75.89%, a PPV of 28.27%, and a NPV of 89.14%.

Overall, the concordance rate between the two methods was 71.9% (Table 3). The majority of concordant cases were double negative, showing high agreement (89.14%) in negative samples, whereas 8% of the events were concordantly double positive (Fig. 1). Among the 373 molecular-positive and culture-negative events, 203 tests were performed during an ongoing fever. Analyzing these false-positive episodes in a multivariate analysis, no variables were predictors. Among the 346 molecular positive/blood culture negative episodes, the mainly represented bacteria were coagulase-negative Staphylococci (70.52%), Streptococcus spp (6.36%), P. aeruginosa (5.78%), and E. faecium/faecalis (5.49%). Importantly, nine samples resulted positive for Aspergillus fumigatus, generally hardly detected by the cultural assays, in patients with a possible invasive fungal disease (host factor and clinical features detected by CT scan; galactomannan negative) as defined by EORTC/MSG Consensus Group [19]. The false negative (molecular negative/blood culture positive) events were mostly represented by coagulase-negative Staphylococci (60.14%) detected in a single blood culture bottle, and therefore of limited clinical relevance. Other potential contaminants not directly included in the SF panel were Micrococcus luteus and Bacillus massiliensis. However, a substantial number of discordances was observed also for clinically more relevant pathogens, including E. coli in 11.89%, and by E. faecium/faecalis in 8.39% of the episodes.

Table 3 Comparison between SeptiFast and blood culture results

We performed direct logistic regression to determine the influence of different factors on the possibility to have concordant results between the two tests. We separately analyzed nine independent variables (underlying disease, neutropenia grade, ongoing antimicrobial therapy, test timing, and different sites of infection). The strongest predictors of concordance between the two methods were the abdomen involvement (OR 2.99, CI 1.56–5,72, p = 0.001), and the presence of grade IV neutropenia, with an OR of 1.59 (CI 1.26–2.03, p < 0.001). We also documented a significant association between concordance and test execution at fever onset (p = 0.041), though the effect is modest (OR 1.27, CI 1.01–1.60). Similarly, a slight increase in the likelihood of having concordant results can be seen in case of ongoing prophylaxis during the tests execution (OR 1.37, CI 1.07–1.77, p = 0.014). After analyzing the factors influencing the overall concordance between tests, we sought to analyze the settings in which the molecular result could be regarded as reliable. Accordingly, we separately analyzed the double positive and double negative tests. We considered the ability of the previously identified factors to influence the concordance with a double positive test result. In order to do this, we excluded from the analysis the double negative tests. We found that both testing during an ongoing fever and while receiving therapy increased the chances of the tests being concordant (OR 2.1 and 2.05, CI 1.39–3.16 and 1.38–3.06, respectively, p < 0.001 for both). We then considered the possible influence on the concordance with a double negative test result. Thus, we excluded from the analysis the double positive tests. We found that testing a patient in antimicrobial prophylaxis (OR 1.44, CI 1.13–1.84, p = 0.003), testing at the onset (OR 1.31, CI 1.04–1.66, p = 0.22), abdominal involvement (OR 1.69, CI 1.14–2.52, p = 0.01), and grade IV neutropenia (OR 1.87, CI 1.50–2.33, p < 0.001) all increased the chances of concordance.

In 26 out of the 514 consecutive patients (5%), infections represent a primary cause of death (IRM). Infection-related deaths included 13 cases of pneumonia, 2 viral and 2 fungal infections. Septic shock was documented in 6/26 cases; SeptiFast resulted positive in all samples (2 Pseudomonas aeruginosa, 2 Klebsiella pneumoniae, 1 Escherichia coli, 1 Stenotrophomonas maltophilia), while concomitant BC in only three of them.


Prompt diagnosis is fundamental to prevent infections. Recently, a growing incidence of Gram-negative rods was reported in most cancer centres [20,21,22,23,24]. The potential adjuvant role of molecular assays in the diagnostic algorithm of infection is a matter of intense debate and of great interest to the clinical community [6, 9].

We analysed the clinical value of SeptiFast in a large cohort of patients, evaluating the settings in which its performance is more reliable.

This analysis confirmed the capacity of a molecular test to detect pathogens in sepsis, since SeptiFast gave positive results in 28.3% of cases, whereas only 17.4% of total samples were culture positive. We confirmed also a high episode-to episode agreement (71.9%) between SeptiFast and blood culture with a higher specificity (75.89%) than sensitivity (50.69%). Other groups have reported similar results, as expected from a molecular test [25].

The choice to compare SeptiFast to BC, still accepted as the standard for the detection of bloodstream infections, could be a limitation, particularly in the setting of ongoing antimicrobial therapy. Thus, the estimated overall specificity of SeptiFast is high but still limited by the comparison with the low accuracy of blood culture in detecting such events. In fact, reflecting on the hypothetically greater diagnostic power of SeptiFast, the above-called false positive could also represent a true microbial identification not elicited by the blood culture. However, in the absence of a molecular-to-molecular test comparison, the presence of false positive molecular tests may result in inappropriate modification to the initial empirical regimen and over-treatment in some episodes. The molecular approach could be particularly helpful for the diagnosis of polymicrobial infections or in case of persistent fever, when the patient has already received antibiotics without any microbial detection. A comparison with another PCR-based technique could be more reliable because of its major sensitivity, the presumed characteristic of molecular approaches.

The presence of false-negative PCR results (in presence of positive BC) could be explained in two ways. Firstly, the bacteria detected by the blood cultures could be contaminant agents that were not present in the SeptiFast sample. In fact, unfortunately because of the retrospective nature of the study, we were unable to ascertain where the SeptiFast sample was taken, and in particular whether it originated from the same site of collection of positive cultures. Therefore, the possibility of a contamination of the blood culture sample by opportunistic microorganisms in some cases still cannot be excluded.

A negative SeptiFast could result in the under-treatment of some cases, determining discontinuance of antibacterial therapy in the setting of febrile neutropenia, underlying the need to associate standard blood cultures to the molecular assay. The incorporation of a wider panel of pathogens within a molecular test may be expected to reach greater diagnostic sensitivity [26] (fewer false negative results), providing a greater confidence in case of a negative test result. Moreover, a majority of Gram-positive bacteria (coagulase-negative staphylococci) were detected in the cases positive only by BC, helping clinicians to differentiate a sample contamination from the isolation of real pathogens.

In our analysis, we found SeptiFast to be more concordant with blood culture in case of absolute neutropenia, during prophylaxis, at fever onset and if abdomen was involved as site of infection. Examining more in depth our data, we concluded that, in the setting just described, only negative SeptiFast resulted to be more concordant with blood culture. In contrast, concordant positive results were more likely to be related to the ongoing fever and antimicrobial therapy. These results were strongly influenced by the double negative cases (1174); double positive ones were only 147.

Interestingly, the clinical setting associated to either double positive or negative results can give suggestions on whether SeptiFast is worthy to be executed in a particular patient on the basis of its reliability. Moreover, even the outpatient management can be influenced by the execution of this test. The time advantage was extremely important for negative samples, as in the case of a patient presenting at the onset of febrile neutropenia during prophylaxis, when SeptiFast negativity could allow doctors to be sure at about 70% not to admit the patient because of the expected concordant negativity in blood culture.

The value of this molecular approach is the rapidity of processing (6 h from the beginning of the extraction phase [8]), remarkable even if compared to the new fast blood cultures, which take at least 48–72 h. This data in the setting already described can compensate for the higher costs [27, 28]. In fact, for example, the expense can be balanced by a hospital admission sparing, as previously described. In addition, in critical patients, in whom a prompt specific therapy is fundamental (particularly during an ongoing fever and prolonged antimicrobial therapy scenario, in which a positive concordance between the two methods is significant), SeptiFast could help in early sparing useless and/or potentially more toxic antibiotics. Recently, Mancini et al. conducted an observational propensity score-matched study to evaluate the costs and clinical outcomes of episodes of suspected sepsis in haematological patients, comparing a retrospective cohort managed with standard assays and a prospective cohort managed with the addition of a molecular assay. Interesting findings were observed in the prospective group, such as a reduction of the costs for antifungal therapy, favouring a more rational use of economic resources [19]. Anyway, the clinical context may impact the ultimate management plan more than a laboratory test, requiring a prospective trial to better address this issue. The retrospective nature of this study prevent us from drawing a conclusion on the value of adding this molecular approach in patient management.

Another aspect that is worth underlining is the SeptiFast ability in detection of fungi (2.5% in our study). In fact, some fungal species (e.g., Aspergillus fumigatus) are not recognised by standard blood culture, and can be consequently difficult to detect. Importantly SetiFast was able to detect Aspergillus spp even in case of serial negative galactomannan antigen tests, thus allowing an early diagnosis, promptly established targeted therapy and severe fungal infection control [29].

The study underlines the potential of molecular-based assays directly performed on blood samples, especially if implementing the detection of antibiotic resistance genes, which is lacking in the used study.


  1. 1.

    Gonsalves MD, Sakr Y. Early identification of sepsis. Curr Infect Dis Rep. 2010;12:329–35.

  2. 2.

    Angus DC, van der Poll T. Severe sepsis and septic shock. N Engl J Med. 2013;369:840–51.

  3. 3.

    Gooley TA, Chien JW, Pergam SA, Hingorani S, Sorror ML, Boeckh M, et al. Reduced mortality after allogeneic hematopoietic-cell transplantation. N Engl J Med. 2010;363:2091–101.

  4. 4.

    Morrissey CO, Chen SC, Sorrell TC, Milliken S, Bardy PG, Bradstock KF, et al. Galactomannan and PCR versus culture and histology for directing use of antifungal treatment for invasive aspergillosis in high-risk haematology patients: a randomised controlled trial. Lancet Infect Dis. 2013;13:519–28.

  5. 5.

    Gratwohl A, Brand R, Frassoni F, Rocha V, Niederwieser D, Reusser P, et al. Cause of death after allogeneic haematopoietic stem cell transplantation (HSCT) in early leukaemias: an EBMT analysis of lethal infectious complications and changes over calendar time. Bone Marrow Transplant. 2005;36:757–69.

  6. 6.

    Mancini N, Carletti S, Ghidoli N, Cichero P, Burioni R, Clementi M. The era of molecular and other non-culture-based methods in diagnosis of sepsis. Clin Microbiol Rev. 2010;23:235–51.

  7. 7.

    Wallet F, Nseir S, Baumann L, Herwegh S, Sendid B, Boulo M, et al. Preliminary clinical study using a multiplex real-time PCR test for the detection of bacterial and fungal DNA directly in blood. Clin Microbiol Infect. 2010;16:774–9.

  8. 8.

    Mancini N, Clerici D, Diotti R, Perotti M, Ghidoli N, De Marco D, et al. Molecular diagnosis of sepsis in neutropenic patients with haematological malignancies. J Med Microbiol. 2008;57:601–4.

  9. 9.

    Varani S, Stanzani M, Paolucci M, Melchionda F, Castellani G, Nardi L, et al. Diagnosis of bloodstream infections in immunocompromised patients by real-time PCR. J Infect. 2009;58:346–51.

  10. 10.

    Bravo D, Blanquer J, Tormo M, Aguilar G, Borras R, Solano C, et al. Diagnostic accuracy and potential clinical value of the LightCycler SeptiFast assay in the management of bloodstream infections occurring in neutropenic and critically ill patients. Int J Infect Dis. 2011;15:e326–31.

  11. 11.

    Guido M, Quattrocchi M, Zizza A, Pasanisi G, Pavone V, Lobreglio G, et al. Molecular approaches in the diagnosis of sepsis in neutropenic patients with haematological malignances. J Prev Med Hyg. 2012;53:104–8.

  12. 12.

    Reers Y, Idelevich EA, Patkau H, Sauerland MC, Tafelski S, Nachtigall I, et al. Multiplex PCR assay underreports true bloodstream infections with coagulase-negative staphylococci in hematological patients with febrile neutropenia. Diagn Microbiol Infect Dis. 2016;85:413–5.

  13. 13.

    Dark P, Wilson C, Blackwood B, McAuley DF, Perkins GD, McMullan R, et al. Accuracy of LightCycler(R) SeptiFast for the detection and identification of pathogens in the blood of patients with suspected sepsis: a systematic review protocol. BMJ Open. 2012;2:e000392.

  14. 14.

    Lamoth F, Jaton K, Prod’hom G, Senn L, Bille J, Calandra T, et al. Multiplex blood PCR in combination with blood cultures for improvement of microbiological documentation of infection in febrile neutropenia. J Clin Microbiol. 2010;48:3510–6.

  15. 15.

    Zaia J, Baden L, Boeckh MJ, Chakrabarti S, Einsele H, Ljungman P, et al. Viral disease prevention after hematopoietic cell transplantation. Bone Marrow Transplant. 2009;44:471–82.

  16. 16.

    Maertens J, Marchetti O, Herbrecht R, Cornely OA, Fluckiger U, Frere P, et al. European guidelines for antifungal management in leukemia and hematopoietic stem cell transplant recipients: summary of the ECIL 3--2009 update. Bone Marrow Transplant. 2011;46:709–18.

  17. 17.

    Freifeld AG, Bow EJ, Sepkowitz KA, Boeckh MJ, Ito JI, Mullen CA, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the infectious diseases society of America. Clin Infect Dis. 2011;52:e56–93.

  18. 18.

    Young JA, Logan BR, Wu J, Wingard JR, Weisdorf DJ, Mudrick C, et al. Infections after transplantation of bone marrow or peripheral blood stem cells from unrelated donors. Biol Blood Marrow Transplant. 2016;22:359–70.

  19. 19.

    De Pauw B, Walsh TJ, Donnelly JP, Stevens DA, Edwards JE, Calandra T, et al. 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. 2008;46:1813–21.

  20. 20.

    Wisplinghoff H, Seifert H, Wenzel RP, Edmond MB. Current trends in the epidemiology of nosocomial bloodstream infections in patients with hematological malignancies and solid neoplasms in hospitals in the United States. Clin Infect Dis. 2003;36:1103–10.

  21. 21.

    Ruhnke M, Arnold R, Gastmeier P. Infection control issues in patients with haematological malignancies in the era of multidrug-resistant bacteria. Lancet Oncol. 2014;15:e606–19.

  22. 22.

    Balletto E, Mikulska M. Bacterial infections in hematopoietic stem cell transplant recipients. Mediterr J Hematol Infect Dis. 2015;7:e2015045.

  23. 23.

    Klastersky J, Ameye L, Maertens J, Georgala A, Muanza F, Aoun M, et al. Bacteraemia in febrile neutropenic cancer patients. Int J Antimicrob Agents. 2007;30(Suppl 1):S51–9.

  24. 24.

    Girmenia C, Rossolini GM, Piciocchi A, Bertaina A, Pisapia G, Pastore D, et al. Infections by carbapenem-resistant Klebsiella pneumoniae in SCT recipients: a nationwide retrospective survey from Italy. Bone Marrow Transplant. 2015;50:282–8.

  25. 25.

    Dark P, Blackwood B, Gates S, McAuley D, Perkins GD, McMullan R, et al. Accuracy of LightCycler((R)) SeptiFast for the detection and identification of pathogens in the blood of patients with suspected sepsis: a systematic review and meta-analysis. Intensive Care Med. 2015;41:21–33.

  26. 26.

    Lehmann LE, Hunfeld KP, Emrich T, Haberhausen G, Wissing H, Hoeft A, et al. A multiplex real-time PCR assay for rapid detection and differentiation of 25 bacterial and fungal pathogens from whole blood samples. Med Microbiol Immunol. 2008;197:313–24.

  27. 27.

    Mancini N, Sambri V, Corti C, Ghidoli N, Tolomelli G, Paolucci M, et al. Cost-effectiveness of blood culture and a multiplex real-time PCR in hematological patients with suspected sepsis: an observational propensity score-matched study. Expert Rev Mol Diagn. 2014;14:623–32.

  28. 28.

    Lagu T, Rothberg MB, Shieh MS, Pekow PS, Steingrub JS, Lindenauer PK. Hospitalizations, costs, and outcomes of severe sepsis in the United States 2003 to 2007. Crit Care Med. 2012;40:754–61.

  29. 29.

    Greco R, Mancini N, Peccatori J, Cieri N, Vago L, Giglio F, et al. Early molecular diagnosis of aspergillosis in a patient with acute myeloid leukaemia. Heart, lung Vessels. 2014;6:119–24.

Download references


Support was provided by the Italian Ministry of Health, University and Research.

Author contributions

R.G, M.C.B, N.M, M.C, and F.C. conceived the study, analyzed data and wrote the manuscript. All authors carried out the collection of patient information and provided clinical care. All authors read and approved the final manuscript.

Author information

Correspondence to Fabio Ciceri.

Ethics declarations

Conflict of interest

The authors declare that they have no conflict of interest.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Greco, R., Barbanti, M.C., Mancini, N. et al. Adjuvant role of SeptiFast to improve the diagnosis of sepsis in a large cohort of hematological patients. Bone Marrow Transplant 53, 410–416 (2018).

Download citation

Further reading