- Email: [email protected]
Use of a multiplex polymerase chain reaction system for enhanced bloodstream pathogen detection in thoracic transplantation
http://www.jhltonline.org
Use of a multiplex polymerase chain reaction system for enhanced bloodstream pathogen detection in thoracic transplantation Antigoni Chaidaroglou, PhD,a Eleni Manoli, MD,b Ekaterini Marathias, MD,c Aggeliki Gkouziouta, MD,d George Saroglou, MD,e Peter Alivizatos, MD,d and Dimitrios Degiannis, MD, PhDa From the aMolecular Immunopathology and Histocompatibility Laboratory, Onassis Cardiac Surgery Center, Athens, Greece; b Microbiology Laboratory, Onassis Cardiac Surgery Center, Athens, Greece; cIntensive Care Unit, Onassis Cardiac Surgery Center, Athens, Greece; dDepartment of Cardiology, Onassis Cardiac Surgery Center, Athens, Greece; and the eThoracic Transplantation Unit, Onassis Cardiac Surgery Center, Athens, Greece.
KEYWORDS: bloodstream infection; blood PCR; molecular diagnosis; septicemia; bacteremia; fungemia; sepsis
BACKGROUND: Bloodstream infections (BSIs) constitute a frequent post-transplant complication in thoracic allograft recipients, especially during the early post-surgical period when patients are under intense immunosuppression. Thus, early and accurate identification of the responsible pathogens is of critical importance for patient survival. In this study we investigated the potential clinical utility of a multiplex real-time polymerase chain reaction (PCR) technology (SeptiFast; Roche Diagnostics) for the detection of BSIs in a cohort of thoracic allograft recipients. METHODS: Our observational study included analysis of 130 blood samples from 30 thoracic allograft recipients (23 heart and 7 lung) using SeptiFast in parallel with blood culture. Samples were drawn when there were clinical and laboratory signs of BSI. The applied molecular assay has been designed to allow direct detection of a wide panel of Gram-positive and Gram-negative bacteria and fungi in blood samples. RESULTS: Real-time PCR yielded concurrent negative and positive results with blood culture methodology in 113 (86.9%) and 5 (3.9%) samples, respectively, with 100% concordance in species identification. SeptiFast identified microorganisms in 9 (6.9%) additional samples that were negative by blood culture. The combined use of SeptiFast and blood culture during the early post-transplant period (o2 months) significantly increased the number of positive samples detected to 17.9% (14 of 78) from 7.7% (6 of 78) detected by blood culture alone (p o 0.05). SeptiFast results were available, on average, within 6 hours from sample collection. CONCLUSIONS: The PCR-based SeptiFast test is a valuable addition to the traditional blood culture method for rapid etiologic diagnosis of BSIs in thoracic transplant recipients, especially during the early post-transplant period. J Heart Lung Transplant 2013;32:707–713 r 2013 International Society for Heart and Lung Transplantation. All rights reserved.
Despite many advances in the field of thoracic organ transplantation, bloodstream infections (BSIs) remain a Reprint requests: Antigoni Chaidaroglou, PhD, Molecular Immunopathology and Histocompatibility Laboratory, Onassis Cardiac Surgery Center, 356 Sygrou Avenue, Athens 17674, Greece. Telephone: þ30-2109493019. Fax: þ30-210-9493018. E-mail address: [email protected]
common and serious complication in this immunocompromised patient population.1 More than 10% of heart and 25% of lung transplant recipients will develop at least 1 episode of BSI post-transplant, with the majority of episodes (460%) developing during the first 2 months after surgery and most having a bacterial or fungal etiology.2–4 Under the intense immunosuppression protocols applied during the early post-transplant period these BSIs may progress
1053-2498/$ - see front matter r 2013 International Society for Heart and Lung Transplantation. All rights reserved. http://dx.doi.org/10.1016/j.healun.2013.04.014
708
The Journal of Heart and Lung Transplantation, Vol 32, No 7, July 2013
rapidly, and several studies have shown that BSI episodes are associated with increased mortality.2,3,5 Suspected BSIs are initially covered with empirical broad-spectrum antibiotics; however, it has been demonstrated that 40% to 70% of all patients with BSIs are not adequately covered during the empirical period.6,7 Inadequate empirical antimicrobial therapy has been associated with excess mortality and prolonged hospitalization.8,9 It is therefore evident that, for improved graft and patient survival, early and accurate identification of the responsible organism and administration of the appropriate treatment is of utmost importance. The “gold standard” in pathogen detection is blood culture; however, there are certain time and sensitivity limitations associated with the method. Results are available no earlier than 24 to 36 hours after sample collection for common bacteria and 41 week for fungi. Furthermore, false negative results are frequently obtained when antibiotics have been administered and when the pathogens involved require special growth conditions for their characterization.10 The emergence of matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) in the field of clinical diagnostic microbiology has considerably shortened the turnaround time required for bacterial and fungal identification.11 However, the capacity of MALDI-TOF MS to identify microorganisms depends on their previous growth in culture media. Therefore, the time required for attaining appropriate growth and the time needed to remove red blood cells before MALDI-TOF MS analysis considerably extends the time interval from blood collection to final results by at least few hours and certainly has to be added to the 1 to 1.5 hours reported to be sufficient to conduct MALDI-TOF MS analysis.11 In addition, MALDI-TOF MS cannot altogether be executed in those cases in which bacterial/fungal culture growth has been challenged by the anti-microbial treatment patients may have already received. The need for an early and rapid detection system for bacterial and fungal BSIs has been addressed by the use of molecular methodologies, which have greater sensitivity than blood cultures and the capability to detect pathogen’s genetic material even in patients receiving antibiotics.10 A newly established multiplex real-time polymerase chain reaction (PCR) assay, termed SeptiFast (Roche Diagnostics, Mannheim, Germany), enables direct detection and identification of the most common bacteria and fungi in blood samples in about 4.5 hours.12,13 In this study we investigated the potential clinical utility of SeptiFast, supplementary to blood culture, for the detection of bacterial and fungal bloodstream infections in heart or lung transplant recipients with suspected BSI.
2008 to November 2010). Samples for blood cultures or PCR were drawn by venipuncture by trained nurses or physicians, using an aseptic technique, which included wearing sterile gloves, a face mask and disposable cap and gown. Samples were drawn from patients suspected of having sepsis due to bacterial or fungal infection based on clinical and laboratory data, and an episode of anti-microbial treatment, and for whom a blood culture was considered appropriate for identification of the causative pathogen. Sepsis was defined according to the consensus guidelines of the American College of Chest Physicians/Society of Critical Care Medicine.14 An episode of anti-microbial treatment was defined as initiation, extension or change of an anti-microbial treatment, with the limitation that it was kept unchanged for ≥72 hours.13 In this context, inclusion of multiple episodes per patient was possible. Samples were drawn and analyzed by SeptiFast only at the discretion of the attending physician and only when a blood culture would normally have been requested, either for identifying the etiologic factor of probable BSI or for monitoring the effectiveness of the antibiotic regime. Consequently, every sample drawn and analyzed by SeptiFast was paired with a blood sample drawn from the same patient, on the same date and at the same time, and analyzed by blood culture. In evaluating our results, the samples analyzed were categorized into 2 clinical time-frames: 78 samples drawn from patients during their early post-transplant period (0 to 60 days, median 18 days) and 52 samples drawn from recipients during their late posttransplant period (61 to 4,641 days, median 165 days). Culture results of samples taken from other body sites (surgical site, lower respiratory track specimens, etc.) on the same date were also evaluated. Furthermore, white blood cell count (WBC), highsensitivity C-reactive protein (hs-CRP) and pro-calcitonin (PCT) levels obtained on corresponding dates were also recorded. BSI was determined on the basis of microbiologic results and physicians’ findings according to definitions established by the Centers for Disease Control and Prevention (CDC).15 Surveillance of BSIs and nosocomial infections was part of the standard practice of care, also as defined by the CDC.16 The study was approved by our hospital’s ethics committee.
Immmunosuppressive therapy All heart or lung transplant recipients received induction immunosuppression with 3 to 5 doses of anti-thymocyte globulin (ATG) and 1 g of intravenous (IV) methylprednisolone while on cardiopulmonary bypass. The initial post-operative imunosuppressive regimen continued with methylprednisolone (125 mg IV every 8 hours for 3 doses, which was quickly tapered over 3 to 4 days to oral prednisone 20 mg/day), mycophenolate mofetil (1 to 2 g/day) and cyclosporine (usually started on the third post-operative day and adjusted so as to maintain an initial level of 250 to 350 ng/ml). Rejection episodes were diagnosed by either endomyocardial or transbronchial biopsies and treated with 1 g of an IV bolus of methylprednisolone for 3 consecutive days. Anti-thymocyte globulin was occasionally administered with due caution for steroid-refractory acute rejection.
Methods Peri-operative anti-microbial prophylaxis Patients and samples A total of 130 peripheral blood samples were collected from 30 heart or lung transplant recipients (23 heart and 7 lung) at the Onassis Cardiac Surgery Center over a period of 2.5 years (May
The same pre-operative protocol was used in both heart and lung transplant recipients. Anti-microbial prophylaxis with ceftazidime (2 g IV every 8 hours) and teichoplanin (800 mg IV every 24 hours) was started on induction of anesthesia and continued for 48
Chaidaroglou et al.
Multiplex PCR for Pathogen Detection
to 72 hours, modified according to the microbiologic data from the donor or the recipient. Anti-fungal prophylaxis was reserved for patients at risk for invasive fungal infections.
709 Table 1 Panel of Microorganisms Detectable by the Multiplex Polymerase Chain Reaction System (SeptiFast) Gram(−)
Gram(þ)
PCT and hs-CRP Measurements
Escherichia coli
PCT and hs-CRP levels were detected using an automated analyzer (COBAS 6000; Roche Diagnostics GmbH, Mannheim, Germany) with reagents, controls and protocols also provided by Roche.
Klebsiella (pneumoniae/oxytoca) Serratia marcescens
Blood culture analysis
Enterobacter (cloacae/aerogenes) Proteus mirabilis
Staphylococcus Candida aureus albicans CoNSa Candida tropicalis Streptococcus Candida pneumoniae parapsilosis Streptococcus Candida sppb glabrata Enterococcus Candida faecium krusei Enterococcus Aspergillus faecalis fumigatus
Blood cultures were performed using an automated blood culture system (Bactec 9120; Becton Dickinson, Heidelberg, Germany) in accordance with the Clinical and Laboratory Standards Institute (CLSI) protocol. Blood cultures were incubated for up to 7 days. Positive blood cultures were considered as contaminated after accounting for the number of independent positive and negative cultures, the probability the organism is a skin contaminant, the other concurrent microbiology results, as well as compatibility with the patient’s clinical status.17 Conventional bacterial/fungal cultures taken from other body sites were performed according to standard microbiologic procedures.
Multiplex real-time PCR (SeptiFast) analysis After blood culture samples were drawn, whole blood was collected under aseptic conditions into an ethylene-diamine tetraacetic acid (EDTA) tube for multiplex PCR testing and then processed normally within 2 hours. Specimen preparation, DNA extraction, PCR amplification and detection of PCR products were performed according to the instructions of the kit manufacturer (Roche Diagnostics). Identification of microorganisms was based on melting curve analysis. Data evaluation was performed by the manual mode of results analysis and subsequently interpreted by the specifically designed SeptiFast Identification Software (SIS; Roche Diagnostics), which automatically calculated the melting peak (Tm) values and the corresponding peak heights. Table 1 lists the organisms detected by the SeptiFast test. According to the manufacturer, the limit of detection was 30 colony-forming units (CFU)/ml for all species, with the exception of Staphylococcus epidermidis, Staphylococcus haemolyticus, Streptococcus agalactiae, Streptococcus pyogenes, Streptococcus pneumoniae and Streptococcus mitis, for which the limit was 100 CFU/ml. For CoNS and Streptococcus species, the SIS analysis program, besides reporting Tm values and peak heights, also reports the semi-quantitative parameter amplification curve crossing point (Cp). The Cp represents the specific amplification cycle at which the concentration of amplified product is above the detection threshold, and is related to the initial concentration of the microorganism in the blood sample. The lower the initial concentration of the microorganism in the sample, the more amplification cycles will be required for the amplicons to exceed the detection threshold, and thus the higher the Cp value will be. To reduce false positive results for CoNS and Streptococcus spp, and based on the assumption that if the initial concentration of these commensal organisms in the blood sample is low then they are contaminants and not causal agents for real infection, the SIS analysis software automatically applies a Cp cut-off threshold, which has been set by the manufacturer of the test at 20 cycles. Cp o20 represents a positive result, whereas a CP 420 is considered a contamination and is displayed as a negative result on the SIS report. However, in the manual mode of result analysis, the actual
Pseudomonas aeruginosa
Fungi
Acinetobacter baumanni Stenotrophomonas maltophilia a Coagulase-negative staphylococci (CoNS) detected as a group: S hominis subsp novobiosepticus, S pasteuri, S wameri, S cohnii subsp urealyticum, S hominis subsp hominis, S lugdunensis, S cohnii subsp cohnii, S capitis subsp ureolyticus, S capitis subsp capitis, S caprae, S saprophyticus, S saprophyticus subsp saprophyticus, S xylosus, S epidermidis, S haemolyticus. b Streptococcal pathogens detected as a group under the name Streptococcus spp: S agalactiae, S anginosus, S bovis, S constellatus, S cristatus, S gordonii, S intermedius, S milleri, S mitis, S mutans, S oralis, S parasanguinis, S pneumoniae, S pyrogenes, S salivarius, S sanguinis, S thermophilus, S viridans, and S vestibularis.
Cp values for any CoNS or Streptococcus spp findings are appropriately displayed for each sample.
Statistical analysis Data were analyzed using SPSS for Windows, version 15 (SPSS, Inc., Chicago, IL). Continuous variables are expressed as mean ⫾ SD. McNemar’s test was used for testing the differences between paired proportions. Two-sided p o 0.05 was considered statistically significant. A 2-sample test for equality of proportions was conducted at a significance level of 5% to compare detection of pathogens when multiplex PCR and blood culture results were combined.
Results Demographic and clinical characteristics of heart and lung transplant recipients involved in the study A total of 30 solid-organ transplant recipients (23 heart, 7 lung) were included in the study and contributed 130 samples, when clinical- and laboratory-based indications of bacterial- or fungi-related sepsis justified blood collection. Patients included 24 (80.0%) men and 6 (20.0%) women, with a median age of 42 years (range 20 to 65 years). The mean white blood cell count was 12.2 103 ⫾ 7.2 103/μl (range 4.0 103 to 36.7 103/μl), whereas the mean CRP level was 62.1 ⫾ 63.9 mg/liter (range 1.7 to 370 mg/liter, n ¼ 122) and mean PCT level was 1.3 ⫾ 4.2 ng/ml (range 0.05 to 32.5 ng/ml, n ¼ 65).
710
The Journal of Heart and Lung Transplantation, Vol 32, No 7, July 2013
Among the 130 samples, 77 were collected while patients were in the ICU and 53 while in hospital wards. Among the 78 samples evaluated during the early post-transplant period, 8 samples were collected from 8 patients during their first 72-hour post-transplant period when they were under surgical prophylaxis. In the remaining 70 samples, 3 samples were drawn while patients were not under antimicrobial treatment, and 67 samples were collected while patients were under empirical or targeted antibiotic treatment. Among the 52 samples evaluated in the late posttransplant period, 17 were drawn from patients who were not on antibiotic treatment prior to sample collection and 35 were drawn while patients were on empirical or targeted antibiotic treatment.
Multiplex real-time PCR and blood culture results of all paired blood samples Cumulative results. The PCR results of the 130 blood samples were compared with the results obtained from the conventional blood culture that had been set in parallel at each individual time-point of sampling. Concordance of SeptiFast and blood culture results was established in 90.8% (118 of 130) of the cases with 5 samples testing positive and 113 testing negative with both methods (Table 2). SeptiFast identified the presence of pathogens in 9 additional samples that had negative blood cultures, whereas 3 samples found positive on blood culture tested negative with real-time PCR. Overall, SeptiFast showed a higher rate of positive results than blood culture (10.7% vs 6.2%), but the difference was not statistically significant. The number of positive samples detected by SeptiFast combined with the number of positive samples detected by blood culture increased to 17 of 130 (13.1%) from 8 of 130 (6.1%) with blood culture alone. This difference in the number of positive samples from the combined use of tests compared with that from blood culture alone was statistically significant (p o 0.01). Bacteria/fungi detected by multiplex PCR and blood culture. In the 5 samples that gave positive results both by multiplex PCR and blood culture there was 100% concordance in the microorganisms identified (Table 3). In these 5 cases with concurrent positivity, SeptiFast results were available, on average, 1.5 days earlier than blood culture results. Similarly, the absence of pathogens in the
Table 2 Comparison of Results of the Multiplex PCR System (SeptiFast) and Blood Culture System (BD Bactec 9120) Blood culture
SeptiFast Positive samples (%) Negative samples (%) Total (%)
Positive Negative samples (%) samples (%) 5 (3.9) 3 (2.3) 8 (6.2)
9 (6.9) 113 (86.9) 122 (93.8)
Total (%) 14 (10.7) 116 (89.3) 130 (100)
blood sample was confirmed considerably earlier with SeptiFast than by blood culture (in o6 hours). Bacteria/fungi detected by multiplex PCR but not by blood culture. Nine samples were characterized as positive by multiplex PCR and negative by blood culture, 1 of which was identified with double infection caused by Candida albicans and Serratia marcescens (Table 3). Seven of the 9 discordant cases were considered to be true positive as confirmed by the presence of clinical signs of infection, microbiologic test results from additional specimens, or positive outcome of PCR-guided antibiotic treatment. Specifically, SeptiFast demonstrated the presence of Pseudomonas aeruginosa in 3 samples from the same patient collected on Days 15, 19 and 34 post-transplant during 3 different episodes of anti-microbial treatment. Although the microorganism was not detected in parallel blood cultures, its presence was confirmed by conventional cultures in surgical wound samples and in pharyngeal swabs. There was also a case of Enterobacter cloacae/ aerogenes, which was clinically relevant as judged by the presence of clinical signs of infection and the fact that symptoms resolved after targeted antibiotic therapy. In addition, there were 2 cases of Candida albicans (from the same patient on different episodes) and 1 case of Candida parapsilosis, where PCR results were verified either by subsequent blood cultures or supported by the fact that symptoms resolved when these patients were treated with targeted anti-fungal treatment. In the remaining 2 cases, where microorganisms were detected by multiplex PCR, the results were not confirmed either by blood culture or by microbiologic tests of various body fluids. Laboratory data provided additional support that the PCR-only positive findings most probably reflected true infection. For the 9 positive only by PCR samples the mean white blood cell count was 18.0 103/μl (range 8.2 103 to 34.0 103/μl), mean CRP value was 86 mg/liter (range 12 to 176 mg/liter) and mean PCT value was 3.2 ng/ml (range 0.08 to 8.5 ng/ ml), whereas the corresponding values for these parameters in the blood culture/PCR-negative group were 11.4 103/μl (range 1.6 103 to 36.7 103/μl) (p o 0.05), 57 mg/liter (range 0.2 to 282 mg/liter) (p o 0.05) and 0.7 ng/ml (range 0.05 to 6.8 ng/ml) (p o 0.05). Bacteria/fungi detected by blood culture but not by multiplex PCR. Diverging results were obtained from 3 samples, 1 positive for Candida albicans and 2 for Staphylococcus epidermidis, where blood cultures were positive and SeptiFast was negative (Table 3). One of the 2 Staphylococcus epidermidis cases was considered to be contamination, because only 1 of the 3 cultures was positive. The second case of Staphylococcus epidermidis was considered a real infection as judged by review of the patient’s records and by the fact that fever resolved only after anti-staphyloccocal treatment was initiated. Interestingly, the corresponding PCR findings, in the manual mode of analysis, were CoNS with Cp 25.8 for the case that was considered blood culture contamination, and CoNS with Cp 21.8 for the case that was considered
Chaidaroglou et al. Table 3
Multiplex PCR for Pathogen Detection
711
Test Results of All Paired Blood Samples Blood culture positive
Blood culture negative
SeptiFast positive
Klebsiella pneumoniae n¼1 Serratia marcescens n¼1 Candida parapsilosis n¼1 Stenotrophomonas maltophilia n¼2 -------------------------------------Total 5
Staphylococcus aureus n¼1 Enterobacter cloacae/aerogenes n¼1 Esherichia coli n¼1 Pseudomonas aeruginosa n¼3 Serratia marcescens n ¼ 1a Candida albicans n ¼ 2a Candida parapsilosis n¼1 ------------------------------------Total 9
SeptiFast negative
Candida albicans Staphylococcus epidermidis ------------------------------Total
n¼1 n ¼ 2b ------3
a Microorganisms b
were detected in the same blood sample. One case of possible contamination.
true infection. The case of Candida albicans, which was detected by blood culture and not by PCR, was also found to be clinically relevant on the basis of clinical signs and response to targeted anti-fungal therapy. Distribution of microorganisms identified by multiplex PCR or/and blood culture during early (0 to 60 days) and late (3 to 155 months) post-transplant periods. During the early post-transplant period (up to 60 days posttransplant), 5 of the 78 samples tested were positive by both PCR and blood culture, whereas 8 additional samples were identified as positive only by PCR (in 1 sample 2 microorganisms were detected) (Table 4). Another sample tested positive by blood culture, whereas PCR failed to detect any pathogen. In the late post-transplant period (3 to 155 months) the rate of positivity among the 52 collected samples was markedly lower than in the early posttransplant period. PCR identified a microorganism in 1 of the 52 samples tested, whereas blood culture detected pathogens in 2 additional samples, reaching a combined positivity of 5.7% (Table 4). During the early post-transplant period, the combined use of blood culture and PCR increased the number of positive samples detected to 17.9% (14 of 78) from 16.7% (13 of 78) that was with SeptiFast alone or from 7.7% (6 of 78) that was with blood culture alone. This difference in the number of positive samples detected from the combined use of tests compared with that from blood culture was statistically significant (p o 0.05).
Discussion During the early post-transplant period, thoracic organ recipients carry the added risk of rapidly progressing BSI due to their immunosuppressive status.4,5,18,19 Therefore, a key factor in minimizing the detrimental effects of a BSI in thoracic organ recipients and improving patients’ prognosis is early detection and identification of the causative
pathogen to expedite administration of the appropriate anti-microbial therapy. The adoption of the SeptiFast methodology to our cohort of patients, led to the detection of significantly more pathogens than blood culture analysis alone, in agreement with results from previous studies.20–23 One possible explanation why a pathogen was not detected in some of these samples by blood culture analysis may be that culture growth might have been affected by antibiotic administration. Actually, 4 of the 9 PCR-positive but bloodculture-negative samples originated from patients receiving antibiotics targeting the pathogen identified by SeptiFast. Among these 9 PCR-only positive samples there were also 3 samples obtained from patients not adequately covered at the time of PCR reporting. When the initial anti-microbial therapy was modified according to PCR results, symptoms were resolved in all of them. Furthermore, additional confirmation for the involvement of the detected by the multiplex PCR assay microorganism in an ongoing BSI was provided when the PCR result obtained from a peripheral blood sample yielded the same pathogen as cultures of other body fluids from the same patient.20,21 This was in fact the case the 3 PCR-only Pseudomonas aeruginosa-positive samples, which yielded this specific microorganism in several different clinical specimens. An additional indication that the PCR-only positive results most probably represented true infection was the fact that the corresponding samples were characterized by significantly elevated WBC count and PCT and CRP levels, which have been shown to be associated with bacteremia or fungemia. In 3 samples of this study, pathogens were only detected by blood culture. In 2 of these samples, blood culture identified Staphylococcus epidermidis, 1 of which was considered to be a case of contamination and the other considered clinically relevant. The fact that the blood culture result was unmatched by PCR could be explained by an adjustment for reduced sensitivity for coagulase-negative staphylococci and streptococci detection by SeptiFast,
712
The Journal of Heart and Lung Transplantation, Vol 32, No 7, July 2013
Table 4 Microorganisms Identified by Multiplex PCR (SeptiFast) and/or BC During Early (0 to 60 Days) and Late Post-transplant Period (3 to 155 Months) Early post-transplantation (n ¼ 78)
Gram (þ) Gram () Fungi Total number of positive samples (%)
PCR only
PCR þ BC
6a 3a 8 (10.2%)
4 1 5 (6.4%)
Late post-transplantation (n ¼ 52)
BC only
PCR only
BC only
1
1
1
1 (1.3%)
1 (1.9%)
1 2 (3.8%)
BC, blood culture; PCR, polymerase chain reaction. a Two microorganisms were detected in the same blood sample.
which has been endorsed by the assay manufacturer. According to this threshold, only CoNS peaks with a Cp of o20 are reported by the SIS analysis program to exclude possible contamination. Application of the Cp cut-off rule may occasionally be at the expense of the sensitivity for these microorganisms. The possibility exists that, occasionally, CoNS or Streptococcus spp amplicons detected with a Cp of slightly 420 (i.e., 20.5 or 21) represent true pathogens and not contaminants. In such instances, although the finding is not formally reported by the SIS analysis program, it may be useful to communicate it to the attending physician along with the information that the detected concentration of CoNS or Streptococcus spp aplicons is marginally under the detection cut-off. Taking into account multiplex PCR results, blood culture results, clinical data and laboratory findings may facilitate the discrimination of CoNS or Streptococcus spp findings as contaminants or true pathogens. In the third sample that was positive by blood culture and negative by multiplex PCR, Candida albicans was identified and, although this pathogen is among the species detected by SeptiFast, the assay failed to detect it. The reason why it could not be detected may lie on a possible genetic variability or mutation of the DNA target that affects the binding of primers or probes in the molecular assay.12 In our cohort of patients, a notably higher number of Candida amplicons were detected by PCR as compared with blood culture. Invasive fungal infections are frequently associated with high morbidity and mortality, particularly in immunocompromised patients and prompt initiation of antifungal therapy is of critical importance in such patients.24 Because early diagnosis of fungal infections by blood culture is often difficult due to the long incubation times required and the lack of sensitivity, rapid and accurate fungi detection using SeptiFast and the timely initiation of appropriate therapy could greatly improve patient outcome. The data presented (see Table 4) show that SeptiFast analysis detected more pathogens than blood culture analysis during the early post-transplant period, when prophylactic or targeted chemotherapy was administered to recipients on maximal immunosuppression. Under these circumstances, it appears that the rate of pathogen detection by blood culture was decreased by the concomitant antimicrobial therapy, whereas the multiplex PCR performance remained relatively unaffected. More importantly, during
the early post-transplant period, pathogen detection increased significantly when PCR was used in conjunction with blood culture relative to the use of blood culture alone. These data suggest that SeptiFast is particularly useful clinically for the identification of pathogens in thoracic organ recipients during the early post-transplant period when they are receiving prophylactic or targeted antibiotic treatment. The rapid result generation by SeptiFast constitutes a very significant time advantage over blood culture both for negative and positive samples. One cannot overlook its higher sensitivity, especially in recipients who are under antibiotic treatment and in those with a blood-borne fungal infection. If both methods yield positive results, SeptiFast has the advantage of providing an answer within 6 hours and recipients can benefit by timely administration of targeted antibiotic therapy. In thoracic allograft recipients with uncomplicated post-transplant recovery, the use of SeptiFast with its sensitive and rapid results could have therapeutic impact by de-escalating broad-spectrum antibiotic treatment and therefore reducing the respective costs and emergence of resistant microorganisms. It should be pointed out, however, that a negative result by real-time PCR cannot exclude the presence of a BSI and does not support withdrawal of antibiotics in clinically suspected septicemia. Even a sensitive PCR-based assay system may occasionally fail to detect the allegedly responsible microorganism in severely septic patients.20,22,25 Another reason one should be cautious in interpreting negative multiplex PCR results is that the SeptiFast detection list does not cover all BSI-causing pathogens. This, besides emphasizing the need for a combined use of SeptiFast and blood culture, also indicates that diagnosis of infection should not rely only on PCR results but should always incorporate any additional clinical and laboratory data available. One limitation of our study is that it was a nonrandomized, single-center study with a small number of cases, which may restrict generalization of results. Despite the small numbers, selection of cases investigated using the molecular methodology was based on clinical needs, and therefore our results may better reflect clinical reality as compared with random use of the test for pre-defined subjects. In conclusion, testing thoracic organ transplant recipients suspected for BSI for the presence of microbes by multiplex PCR improves pathogen detection and may help in
Chaidaroglou et al.
Multiplex PCR for Pathogen Detection
713
optimizing anti-microbial treatment of these severely immunocompromised patients. Use of the molecular methodology is suggested, especially in the early post-transplant period when recipients are at higher risk of infection and are in their most vulnerable post-surgical phase. Its capacity to identify blood-borne pathogens within 6 hours combined with its greater sensitivity justify the use of the multiplex PCR method for earlier diagnosis and prompt treatment of infection with appropriate antibiotics, which could lead to better clinical outcomes. However, because SeptiFast cannot detect a number of pathogens or determine antibiotic susceptibility, it should always be used in conjunction with traditional blood culture methods.
9. Garnacho-Montero J, Ortiz-Leyba C, Herrera-Melero I, et al. Mortality and morbidity attributable to inadequate empirical antimicrobial therapy in patients admitted to the ICU with sepsis: a matched cohort study. J Antimicrob Chemother 2008;61:436-41. 10. Peters RP, van Agtmael MA, Danner SA, et al. New developments in the diagnosis of bloodstream infections. Lancet Infect Dis 2004;4: 751-60. 11. Croxatto A, Prod’hom G, Greub G. Applications of MALDI-TOF mass spectrometry in clinical diagnostic microbiology. FEMS Microbiol Rev 2012;36:380-407. 12. Lehmann LE, Hunfeld KP, Emrich T, et al. A multiplex real-time PCR assay for the rapid detection and differentiation of 25 bacterial and fungal pathogens from whole blood samples. Med Microbiol Immunol 2008;197:313-24. 13. Lehmann LE, Alvarez J, Hunfeld KP, et al. Potential clinical utility of polymerase chain reaction in microbiological testing for sepsis. Crit Care Med 2009;37:3085-90. 14. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med, 1992;20:864-74. 15. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections. Am J Infect Control 1988;16:128-40. 16. National Nosocomial Infections Surveillance System. National Nosocomial Infections Surveillance System report: data summary from January 1992 to June 2002. Am J Infect Control 2002;30:458-75. 17. Hall KK, Lyman JA. Updated review of blood culture contamination. Clin Microbiol Rev 2006;19:788-802. 18. Snydman DR. Epidemiology of infections after solid-organ transplantation. Clin Infect Dis 2001;33 Suppl 1:S5-8. 19. Haddad F, Deuse T, Pham M, et al. Changing trends in infectious disease in heart transplantation. J Heart Lung Transplant 2010;29:306-15. 20. Louie RF, Tang Z, Albertson TE, et al. Multiplex polymerase chain reaction detection enhancement of bacteremia and fungemia. Crit Care Med 2008;36:1487-92. 21. Lehmann LE, Hunfeld KP, Steinbrucker M, et al. Improved detection of blood stream pathogens by real-time PCR in severe sepsis. Intensive Care Med 2010;36:49-56. 22. Bloos F, Hinder F, Becker K, et al. A multicenter trial to compare blood culture with polymerase chain reaction in severe human sepsis. Intensive Care Med 2010;36:241-7. 23. Bravo D, Blanquer J, Tormo M, 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. 24. Marik PE. Fungal infections in solid organ transplantation. Expert Opin Pharmacother 2006;7:297-305. 25. Tsalik EL, Jones D, Nicholson B, et al. Multiplex PCR to diagnose bloodstream infections in patients admitted from the emergency department with sepsis. J Clin Microbiol 2010;48:26-33.
Disclosure statement The authors have no conflicts of interest to disclose. The work was presented in part at the 31st annual meeting and scientific sessions of the International Society for Heart and Lung Transplantation, April 2011, San Diego, California.
References 1. Mattner F, Fischer S, Weissbrodt H, et al. Post-operative nosocomial infections after lung and heart transplantation. J Heart Lung Transplant 2007;26:241-9. 2. Kramer MR, Marshall SE, Starnes VA, et al. Infectious complications in heart–lung transplant. Analysis of 200 episodes. Arch Intern Med 1993;153:2010-6. 3. Palmer SM, Alexander BD, Sanders LL, et al. Significance of bloodstream infection after lung transplantation: analysis of 176 consecutive patients. Transplantation 2000;69:2360-6. 4. van de Beek D, Kremers WK, Del Roso JL, et al. Effect of infectious diseases on outcome after heart transplantation. Mayo Clin Proc 2008;83:304-8. 5. Rodriguez C, Muñoz P, Rodriguez-Créixems M, et al. Bloodstream infections among heart transplant recipients. Transplantation 2006;81:384-91. 6. Ibrahim EH, Sherman G, Ward S, et al. The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest 2000;118:146-55. 7. MacArthur RD, Miller M, Albertson T, et al. Adequacy of early empiric antibiotic treatment and survival in severe sepsis: experience from the MONARCS trial. Clin Infect Dis 2004;38:284-8. 8. Kollef MH, Sherman G, Ward S, et al. Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest 1999;115:462-74.
Use of a multiplex polymerase chain reaction system for enhanced bloodstream pathogen detection in thoracic transplantation Antigoni Chaidaroglou, PhD,a Eleni Manoli, MD,b Ekaterini Marathias, MD,c Aggeliki Gkouziouta, MD,d George Saroglou, MD,e Peter Alivizatos, MD,d and Dimitrios Degiannis, MD, PhDa From the aMolecular Immunopathology and Histocompatibility Laboratory, Onassis Cardiac Surgery Center, Athens, Greece; b Microbiology Laboratory, Onassis Cardiac Surgery Center, Athens, Greece; cIntensive Care Unit, Onassis Cardiac Surgery Center, Athens, Greece; dDepartment of Cardiology, Onassis Cardiac Surgery Center, Athens, Greece; and the eThoracic Transplantation Unit, Onassis Cardiac Surgery Center, Athens, Greece.
KEYWORDS: bloodstream infection; blood PCR; molecular diagnosis; septicemia; bacteremia; fungemia; sepsis
BACKGROUND: Bloodstream infections (BSIs) constitute a frequent post-transplant complication in thoracic allograft recipients, especially during the early post-surgical period when patients are under intense immunosuppression. Thus, early and accurate identification of the responsible pathogens is of critical importance for patient survival. In this study we investigated the potential clinical utility of a multiplex real-time polymerase chain reaction (PCR) technology (SeptiFast; Roche Diagnostics) for the detection of BSIs in a cohort of thoracic allograft recipients. METHODS: Our observational study included analysis of 130 blood samples from 30 thoracic allograft recipients (23 heart and 7 lung) using SeptiFast in parallel with blood culture. Samples were drawn when there were clinical and laboratory signs of BSI. The applied molecular assay has been designed to allow direct detection of a wide panel of Gram-positive and Gram-negative bacteria and fungi in blood samples. RESULTS: Real-time PCR yielded concurrent negative and positive results with blood culture methodology in 113 (86.9%) and 5 (3.9%) samples, respectively, with 100% concordance in species identification. SeptiFast identified microorganisms in 9 (6.9%) additional samples that were negative by blood culture. The combined use of SeptiFast and blood culture during the early post-transplant period (o2 months) significantly increased the number of positive samples detected to 17.9% (14 of 78) from 7.7% (6 of 78) detected by blood culture alone (p o 0.05). SeptiFast results were available, on average, within 6 hours from sample collection. CONCLUSIONS: The PCR-based SeptiFast test is a valuable addition to the traditional blood culture method for rapid etiologic diagnosis of BSIs in thoracic transplant recipients, especially during the early post-transplant period. J Heart Lung Transplant 2013;32:707–713 r 2013 International Society for Heart and Lung Transplantation. All rights reserved.
Despite many advances in the field of thoracic organ transplantation, bloodstream infections (BSIs) remain a Reprint requests: Antigoni Chaidaroglou, PhD, Molecular Immunopathology and Histocompatibility Laboratory, Onassis Cardiac Surgery Center, 356 Sygrou Avenue, Athens 17674, Greece. Telephone: þ30-2109493019. Fax: þ30-210-9493018. E-mail address: [email protected]
common and serious complication in this immunocompromised patient population.1 More than 10% of heart and 25% of lung transplant recipients will develop at least 1 episode of BSI post-transplant, with the majority of episodes (460%) developing during the first 2 months after surgery and most having a bacterial or fungal etiology.2–4 Under the intense immunosuppression protocols applied during the early post-transplant period these BSIs may progress
1053-2498/$ - see front matter r 2013 International Society for Heart and Lung Transplantation. All rights reserved. http://dx.doi.org/10.1016/j.healun.2013.04.014
708
The Journal of Heart and Lung Transplantation, Vol 32, No 7, July 2013
rapidly, and several studies have shown that BSI episodes are associated with increased mortality.2,3,5 Suspected BSIs are initially covered with empirical broad-spectrum antibiotics; however, it has been demonstrated that 40% to 70% of all patients with BSIs are not adequately covered during the empirical period.6,7 Inadequate empirical antimicrobial therapy has been associated with excess mortality and prolonged hospitalization.8,9 It is therefore evident that, for improved graft and patient survival, early and accurate identification of the responsible organism and administration of the appropriate treatment is of utmost importance. The “gold standard” in pathogen detection is blood culture; however, there are certain time and sensitivity limitations associated with the method. Results are available no earlier than 24 to 36 hours after sample collection for common bacteria and 41 week for fungi. Furthermore, false negative results are frequently obtained when antibiotics have been administered and when the pathogens involved require special growth conditions for their characterization.10 The emergence of matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (MS) in the field of clinical diagnostic microbiology has considerably shortened the turnaround time required for bacterial and fungal identification.11 However, the capacity of MALDI-TOF MS to identify microorganisms depends on their previous growth in culture media. Therefore, the time required for attaining appropriate growth and the time needed to remove red blood cells before MALDI-TOF MS analysis considerably extends the time interval from blood collection to final results by at least few hours and certainly has to be added to the 1 to 1.5 hours reported to be sufficient to conduct MALDI-TOF MS analysis.11 In addition, MALDI-TOF MS cannot altogether be executed in those cases in which bacterial/fungal culture growth has been challenged by the anti-microbial treatment patients may have already received. The need for an early and rapid detection system for bacterial and fungal BSIs has been addressed by the use of molecular methodologies, which have greater sensitivity than blood cultures and the capability to detect pathogen’s genetic material even in patients receiving antibiotics.10 A newly established multiplex real-time polymerase chain reaction (PCR) assay, termed SeptiFast (Roche Diagnostics, Mannheim, Germany), enables direct detection and identification of the most common bacteria and fungi in blood samples in about 4.5 hours.12,13 In this study we investigated the potential clinical utility of SeptiFast, supplementary to blood culture, for the detection of bacterial and fungal bloodstream infections in heart or lung transplant recipients with suspected BSI.
2008 to November 2010). Samples for blood cultures or PCR were drawn by venipuncture by trained nurses or physicians, using an aseptic technique, which included wearing sterile gloves, a face mask and disposable cap and gown. Samples were drawn from patients suspected of having sepsis due to bacterial or fungal infection based on clinical and laboratory data, and an episode of anti-microbial treatment, and for whom a blood culture was considered appropriate for identification of the causative pathogen. Sepsis was defined according to the consensus guidelines of the American College of Chest Physicians/Society of Critical Care Medicine.14 An episode of anti-microbial treatment was defined as initiation, extension or change of an anti-microbial treatment, with the limitation that it was kept unchanged for ≥72 hours.13 In this context, inclusion of multiple episodes per patient was possible. Samples were drawn and analyzed by SeptiFast only at the discretion of the attending physician and only when a blood culture would normally have been requested, either for identifying the etiologic factor of probable BSI or for monitoring the effectiveness of the antibiotic regime. Consequently, every sample drawn and analyzed by SeptiFast was paired with a blood sample drawn from the same patient, on the same date and at the same time, and analyzed by blood culture. In evaluating our results, the samples analyzed were categorized into 2 clinical time-frames: 78 samples drawn from patients during their early post-transplant period (0 to 60 days, median 18 days) and 52 samples drawn from recipients during their late posttransplant period (61 to 4,641 days, median 165 days). Culture results of samples taken from other body sites (surgical site, lower respiratory track specimens, etc.) on the same date were also evaluated. Furthermore, white blood cell count (WBC), highsensitivity C-reactive protein (hs-CRP) and pro-calcitonin (PCT) levels obtained on corresponding dates were also recorded. BSI was determined on the basis of microbiologic results and physicians’ findings according to definitions established by the Centers for Disease Control and Prevention (CDC).15 Surveillance of BSIs and nosocomial infections was part of the standard practice of care, also as defined by the CDC.16 The study was approved by our hospital’s ethics committee.
Immmunosuppressive therapy All heart or lung transplant recipients received induction immunosuppression with 3 to 5 doses of anti-thymocyte globulin (ATG) and 1 g of intravenous (IV) methylprednisolone while on cardiopulmonary bypass. The initial post-operative imunosuppressive regimen continued with methylprednisolone (125 mg IV every 8 hours for 3 doses, which was quickly tapered over 3 to 4 days to oral prednisone 20 mg/day), mycophenolate mofetil (1 to 2 g/day) and cyclosporine (usually started on the third post-operative day and adjusted so as to maintain an initial level of 250 to 350 ng/ml). Rejection episodes were diagnosed by either endomyocardial or transbronchial biopsies and treated with 1 g of an IV bolus of methylprednisolone for 3 consecutive days. Anti-thymocyte globulin was occasionally administered with due caution for steroid-refractory acute rejection.
Methods Peri-operative anti-microbial prophylaxis Patients and samples A total of 130 peripheral blood samples were collected from 30 heart or lung transplant recipients (23 heart and 7 lung) at the Onassis Cardiac Surgery Center over a period of 2.5 years (May
The same pre-operative protocol was used in both heart and lung transplant recipients. Anti-microbial prophylaxis with ceftazidime (2 g IV every 8 hours) and teichoplanin (800 mg IV every 24 hours) was started on induction of anesthesia and continued for 48
Chaidaroglou et al.
Multiplex PCR for Pathogen Detection
to 72 hours, modified according to the microbiologic data from the donor or the recipient. Anti-fungal prophylaxis was reserved for patients at risk for invasive fungal infections.
709 Table 1 Panel of Microorganisms Detectable by the Multiplex Polymerase Chain Reaction System (SeptiFast) Gram(−)
Gram(þ)
PCT and hs-CRP Measurements
Escherichia coli
PCT and hs-CRP levels were detected using an automated analyzer (COBAS 6000; Roche Diagnostics GmbH, Mannheim, Germany) with reagents, controls and protocols also provided by Roche.
Klebsiella (pneumoniae/oxytoca) Serratia marcescens
Blood culture analysis
Enterobacter (cloacae/aerogenes) Proteus mirabilis
Staphylococcus Candida aureus albicans CoNSa Candida tropicalis Streptococcus Candida pneumoniae parapsilosis Streptococcus Candida sppb glabrata Enterococcus Candida faecium krusei Enterococcus Aspergillus faecalis fumigatus
Blood cultures were performed using an automated blood culture system (Bactec 9120; Becton Dickinson, Heidelberg, Germany) in accordance with the Clinical and Laboratory Standards Institute (CLSI) protocol. Blood cultures were incubated for up to 7 days. Positive blood cultures were considered as contaminated after accounting for the number of independent positive and negative cultures, the probability the organism is a skin contaminant, the other concurrent microbiology results, as well as compatibility with the patient’s clinical status.17 Conventional bacterial/fungal cultures taken from other body sites were performed according to standard microbiologic procedures.
Multiplex real-time PCR (SeptiFast) analysis After blood culture samples were drawn, whole blood was collected under aseptic conditions into an ethylene-diamine tetraacetic acid (EDTA) tube for multiplex PCR testing and then processed normally within 2 hours. Specimen preparation, DNA extraction, PCR amplification and detection of PCR products were performed according to the instructions of the kit manufacturer (Roche Diagnostics). Identification of microorganisms was based on melting curve analysis. Data evaluation was performed by the manual mode of results analysis and subsequently interpreted by the specifically designed SeptiFast Identification Software (SIS; Roche Diagnostics), which automatically calculated the melting peak (Tm) values and the corresponding peak heights. Table 1 lists the organisms detected by the SeptiFast test. According to the manufacturer, the limit of detection was 30 colony-forming units (CFU)/ml for all species, with the exception of Staphylococcus epidermidis, Staphylococcus haemolyticus, Streptococcus agalactiae, Streptococcus pyogenes, Streptococcus pneumoniae and Streptococcus mitis, for which the limit was 100 CFU/ml. For CoNS and Streptococcus species, the SIS analysis program, besides reporting Tm values and peak heights, also reports the semi-quantitative parameter amplification curve crossing point (Cp). The Cp represents the specific amplification cycle at which the concentration of amplified product is above the detection threshold, and is related to the initial concentration of the microorganism in the blood sample. The lower the initial concentration of the microorganism in the sample, the more amplification cycles will be required for the amplicons to exceed the detection threshold, and thus the higher the Cp value will be. To reduce false positive results for CoNS and Streptococcus spp, and based on the assumption that if the initial concentration of these commensal organisms in the blood sample is low then they are contaminants and not causal agents for real infection, the SIS analysis software automatically applies a Cp cut-off threshold, which has been set by the manufacturer of the test at 20 cycles. Cp o20 represents a positive result, whereas a CP 420 is considered a contamination and is displayed as a negative result on the SIS report. However, in the manual mode of result analysis, the actual
Pseudomonas aeruginosa
Fungi
Acinetobacter baumanni Stenotrophomonas maltophilia a Coagulase-negative staphylococci (CoNS) detected as a group: S hominis subsp novobiosepticus, S pasteuri, S wameri, S cohnii subsp urealyticum, S hominis subsp hominis, S lugdunensis, S cohnii subsp cohnii, S capitis subsp ureolyticus, S capitis subsp capitis, S caprae, S saprophyticus, S saprophyticus subsp saprophyticus, S xylosus, S epidermidis, S haemolyticus. b Streptococcal pathogens detected as a group under the name Streptococcus spp: S agalactiae, S anginosus, S bovis, S constellatus, S cristatus, S gordonii, S intermedius, S milleri, S mitis, S mutans, S oralis, S parasanguinis, S pneumoniae, S pyrogenes, S salivarius, S sanguinis, S thermophilus, S viridans, and S vestibularis.
Cp values for any CoNS or Streptococcus spp findings are appropriately displayed for each sample.
Statistical analysis Data were analyzed using SPSS for Windows, version 15 (SPSS, Inc., Chicago, IL). Continuous variables are expressed as mean ⫾ SD. McNemar’s test was used for testing the differences between paired proportions. Two-sided p o 0.05 was considered statistically significant. A 2-sample test for equality of proportions was conducted at a significance level of 5% to compare detection of pathogens when multiplex PCR and blood culture results were combined.
Results Demographic and clinical characteristics of heart and lung transplant recipients involved in the study A total of 30 solid-organ transplant recipients (23 heart, 7 lung) were included in the study and contributed 130 samples, when clinical- and laboratory-based indications of bacterial- or fungi-related sepsis justified blood collection. Patients included 24 (80.0%) men and 6 (20.0%) women, with a median age of 42 years (range 20 to 65 years). The mean white blood cell count was 12.2 103 ⫾ 7.2 103/μl (range 4.0 103 to 36.7 103/μl), whereas the mean CRP level was 62.1 ⫾ 63.9 mg/liter (range 1.7 to 370 mg/liter, n ¼ 122) and mean PCT level was 1.3 ⫾ 4.2 ng/ml (range 0.05 to 32.5 ng/ml, n ¼ 65).
710
The Journal of Heart and Lung Transplantation, Vol 32, No 7, July 2013
Among the 130 samples, 77 were collected while patients were in the ICU and 53 while in hospital wards. Among the 78 samples evaluated during the early post-transplant period, 8 samples were collected from 8 patients during their first 72-hour post-transplant period when they were under surgical prophylaxis. In the remaining 70 samples, 3 samples were drawn while patients were not under antimicrobial treatment, and 67 samples were collected while patients were under empirical or targeted antibiotic treatment. Among the 52 samples evaluated in the late posttransplant period, 17 were drawn from patients who were not on antibiotic treatment prior to sample collection and 35 were drawn while patients were on empirical or targeted antibiotic treatment.
Multiplex real-time PCR and blood culture results of all paired blood samples Cumulative results. The PCR results of the 130 blood samples were compared with the results obtained from the conventional blood culture that had been set in parallel at each individual time-point of sampling. Concordance of SeptiFast and blood culture results was established in 90.8% (118 of 130) of the cases with 5 samples testing positive and 113 testing negative with both methods (Table 2). SeptiFast identified the presence of pathogens in 9 additional samples that had negative blood cultures, whereas 3 samples found positive on blood culture tested negative with real-time PCR. Overall, SeptiFast showed a higher rate of positive results than blood culture (10.7% vs 6.2%), but the difference was not statistically significant. The number of positive samples detected by SeptiFast combined with the number of positive samples detected by blood culture increased to 17 of 130 (13.1%) from 8 of 130 (6.1%) with blood culture alone. This difference in the number of positive samples from the combined use of tests compared with that from blood culture alone was statistically significant (p o 0.01). Bacteria/fungi detected by multiplex PCR and blood culture. In the 5 samples that gave positive results both by multiplex PCR and blood culture there was 100% concordance in the microorganisms identified (Table 3). In these 5 cases with concurrent positivity, SeptiFast results were available, on average, 1.5 days earlier than blood culture results. Similarly, the absence of pathogens in the
Table 2 Comparison of Results of the Multiplex PCR System (SeptiFast) and Blood Culture System (BD Bactec 9120) Blood culture
SeptiFast Positive samples (%) Negative samples (%) Total (%)
Positive Negative samples (%) samples (%) 5 (3.9) 3 (2.3) 8 (6.2)
9 (6.9) 113 (86.9) 122 (93.8)
Total (%) 14 (10.7) 116 (89.3) 130 (100)
blood sample was confirmed considerably earlier with SeptiFast than by blood culture (in o6 hours). Bacteria/fungi detected by multiplex PCR but not by blood culture. Nine samples were characterized as positive by multiplex PCR and negative by blood culture, 1 of which was identified with double infection caused by Candida albicans and Serratia marcescens (Table 3). Seven of the 9 discordant cases were considered to be true positive as confirmed by the presence of clinical signs of infection, microbiologic test results from additional specimens, or positive outcome of PCR-guided antibiotic treatment. Specifically, SeptiFast demonstrated the presence of Pseudomonas aeruginosa in 3 samples from the same patient collected on Days 15, 19 and 34 post-transplant during 3 different episodes of anti-microbial treatment. Although the microorganism was not detected in parallel blood cultures, its presence was confirmed by conventional cultures in surgical wound samples and in pharyngeal swabs. There was also a case of Enterobacter cloacae/ aerogenes, which was clinically relevant as judged by the presence of clinical signs of infection and the fact that symptoms resolved after targeted antibiotic therapy. In addition, there were 2 cases of Candida albicans (from the same patient on different episodes) and 1 case of Candida parapsilosis, where PCR results were verified either by subsequent blood cultures or supported by the fact that symptoms resolved when these patients were treated with targeted anti-fungal treatment. In the remaining 2 cases, where microorganisms were detected by multiplex PCR, the results were not confirmed either by blood culture or by microbiologic tests of various body fluids. Laboratory data provided additional support that the PCR-only positive findings most probably reflected true infection. For the 9 positive only by PCR samples the mean white blood cell count was 18.0 103/μl (range 8.2 103 to 34.0 103/μl), mean CRP value was 86 mg/liter (range 12 to 176 mg/liter) and mean PCT value was 3.2 ng/ml (range 0.08 to 8.5 ng/ ml), whereas the corresponding values for these parameters in the blood culture/PCR-negative group were 11.4 103/μl (range 1.6 103 to 36.7 103/μl) (p o 0.05), 57 mg/liter (range 0.2 to 282 mg/liter) (p o 0.05) and 0.7 ng/ml (range 0.05 to 6.8 ng/ml) (p o 0.05). Bacteria/fungi detected by blood culture but not by multiplex PCR. Diverging results were obtained from 3 samples, 1 positive for Candida albicans and 2 for Staphylococcus epidermidis, where blood cultures were positive and SeptiFast was negative (Table 3). One of the 2 Staphylococcus epidermidis cases was considered to be contamination, because only 1 of the 3 cultures was positive. The second case of Staphylococcus epidermidis was considered a real infection as judged by review of the patient’s records and by the fact that fever resolved only after anti-staphyloccocal treatment was initiated. Interestingly, the corresponding PCR findings, in the manual mode of analysis, were CoNS with Cp 25.8 for the case that was considered blood culture contamination, and CoNS with Cp 21.8 for the case that was considered
Chaidaroglou et al. Table 3
Multiplex PCR for Pathogen Detection
711
Test Results of All Paired Blood Samples Blood culture positive
Blood culture negative
SeptiFast positive
Klebsiella pneumoniae n¼1 Serratia marcescens n¼1 Candida parapsilosis n¼1 Stenotrophomonas maltophilia n¼2 -------------------------------------Total 5
Staphylococcus aureus n¼1 Enterobacter cloacae/aerogenes n¼1 Esherichia coli n¼1 Pseudomonas aeruginosa n¼3 Serratia marcescens n ¼ 1a Candida albicans n ¼ 2a Candida parapsilosis n¼1 ------------------------------------Total 9
SeptiFast negative
Candida albicans Staphylococcus epidermidis ------------------------------Total
n¼1 n ¼ 2b ------3
a Microorganisms b
were detected in the same blood sample. One case of possible contamination.
true infection. The case of Candida albicans, which was detected by blood culture and not by PCR, was also found to be clinically relevant on the basis of clinical signs and response to targeted anti-fungal therapy. Distribution of microorganisms identified by multiplex PCR or/and blood culture during early (0 to 60 days) and late (3 to 155 months) post-transplant periods. During the early post-transplant period (up to 60 days posttransplant), 5 of the 78 samples tested were positive by both PCR and blood culture, whereas 8 additional samples were identified as positive only by PCR (in 1 sample 2 microorganisms were detected) (Table 4). Another sample tested positive by blood culture, whereas PCR failed to detect any pathogen. In the late post-transplant period (3 to 155 months) the rate of positivity among the 52 collected samples was markedly lower than in the early posttransplant period. PCR identified a microorganism in 1 of the 52 samples tested, whereas blood culture detected pathogens in 2 additional samples, reaching a combined positivity of 5.7% (Table 4). During the early post-transplant period, the combined use of blood culture and PCR increased the number of positive samples detected to 17.9% (14 of 78) from 16.7% (13 of 78) that was with SeptiFast alone or from 7.7% (6 of 78) that was with blood culture alone. This difference in the number of positive samples detected from the combined use of tests compared with that from blood culture was statistically significant (p o 0.05).
Discussion During the early post-transplant period, thoracic organ recipients carry the added risk of rapidly progressing BSI due to their immunosuppressive status.4,5,18,19 Therefore, a key factor in minimizing the detrimental effects of a BSI in thoracic organ recipients and improving patients’ prognosis is early detection and identification of the causative
pathogen to expedite administration of the appropriate anti-microbial therapy. The adoption of the SeptiFast methodology to our cohort of patients, led to the detection of significantly more pathogens than blood culture analysis alone, in agreement with results from previous studies.20–23 One possible explanation why a pathogen was not detected in some of these samples by blood culture analysis may be that culture growth might have been affected by antibiotic administration. Actually, 4 of the 9 PCR-positive but bloodculture-negative samples originated from patients receiving antibiotics targeting the pathogen identified by SeptiFast. Among these 9 PCR-only positive samples there were also 3 samples obtained from patients not adequately covered at the time of PCR reporting. When the initial anti-microbial therapy was modified according to PCR results, symptoms were resolved in all of them. Furthermore, additional confirmation for the involvement of the detected by the multiplex PCR assay microorganism in an ongoing BSI was provided when the PCR result obtained from a peripheral blood sample yielded the same pathogen as cultures of other body fluids from the same patient.20,21 This was in fact the case the 3 PCR-only Pseudomonas aeruginosa-positive samples, which yielded this specific microorganism in several different clinical specimens. An additional indication that the PCR-only positive results most probably represented true infection was the fact that the corresponding samples were characterized by significantly elevated WBC count and PCT and CRP levels, which have been shown to be associated with bacteremia or fungemia. In 3 samples of this study, pathogens were only detected by blood culture. In 2 of these samples, blood culture identified Staphylococcus epidermidis, 1 of which was considered to be a case of contamination and the other considered clinically relevant. The fact that the blood culture result was unmatched by PCR could be explained by an adjustment for reduced sensitivity for coagulase-negative staphylococci and streptococci detection by SeptiFast,
712
The Journal of Heart and Lung Transplantation, Vol 32, No 7, July 2013
Table 4 Microorganisms Identified by Multiplex PCR (SeptiFast) and/or BC During Early (0 to 60 Days) and Late Post-transplant Period (3 to 155 Months) Early post-transplantation (n ¼ 78)
Gram (þ) Gram () Fungi Total number of positive samples (%)
PCR only
PCR þ BC
6a 3a 8 (10.2%)
4 1 5 (6.4%)
Late post-transplantation (n ¼ 52)
BC only
PCR only
BC only
1
1
1
1 (1.3%)
1 (1.9%)
1 2 (3.8%)
BC, blood culture; PCR, polymerase chain reaction. a Two microorganisms were detected in the same blood sample.
which has been endorsed by the assay manufacturer. According to this threshold, only CoNS peaks with a Cp of o20 are reported by the SIS analysis program to exclude possible contamination. Application of the Cp cut-off rule may occasionally be at the expense of the sensitivity for these microorganisms. The possibility exists that, occasionally, CoNS or Streptococcus spp amplicons detected with a Cp of slightly 420 (i.e., 20.5 or 21) represent true pathogens and not contaminants. In such instances, although the finding is not formally reported by the SIS analysis program, it may be useful to communicate it to the attending physician along with the information that the detected concentration of CoNS or Streptococcus spp aplicons is marginally under the detection cut-off. Taking into account multiplex PCR results, blood culture results, clinical data and laboratory findings may facilitate the discrimination of CoNS or Streptococcus spp findings as contaminants or true pathogens. In the third sample that was positive by blood culture and negative by multiplex PCR, Candida albicans was identified and, although this pathogen is among the species detected by SeptiFast, the assay failed to detect it. The reason why it could not be detected may lie on a possible genetic variability or mutation of the DNA target that affects the binding of primers or probes in the molecular assay.12 In our cohort of patients, a notably higher number of Candida amplicons were detected by PCR as compared with blood culture. Invasive fungal infections are frequently associated with high morbidity and mortality, particularly in immunocompromised patients and prompt initiation of antifungal therapy is of critical importance in such patients.24 Because early diagnosis of fungal infections by blood culture is often difficult due to the long incubation times required and the lack of sensitivity, rapid and accurate fungi detection using SeptiFast and the timely initiation of appropriate therapy could greatly improve patient outcome. The data presented (see Table 4) show that SeptiFast analysis detected more pathogens than blood culture analysis during the early post-transplant period, when prophylactic or targeted chemotherapy was administered to recipients on maximal immunosuppression. Under these circumstances, it appears that the rate of pathogen detection by blood culture was decreased by the concomitant antimicrobial therapy, whereas the multiplex PCR performance remained relatively unaffected. More importantly, during
the early post-transplant period, pathogen detection increased significantly when PCR was used in conjunction with blood culture relative to the use of blood culture alone. These data suggest that SeptiFast is particularly useful clinically for the identification of pathogens in thoracic organ recipients during the early post-transplant period when they are receiving prophylactic or targeted antibiotic treatment. The rapid result generation by SeptiFast constitutes a very significant time advantage over blood culture both for negative and positive samples. One cannot overlook its higher sensitivity, especially in recipients who are under antibiotic treatment and in those with a blood-borne fungal infection. If both methods yield positive results, SeptiFast has the advantage of providing an answer within 6 hours and recipients can benefit by timely administration of targeted antibiotic therapy. In thoracic allograft recipients with uncomplicated post-transplant recovery, the use of SeptiFast with its sensitive and rapid results could have therapeutic impact by de-escalating broad-spectrum antibiotic treatment and therefore reducing the respective costs and emergence of resistant microorganisms. It should be pointed out, however, that a negative result by real-time PCR cannot exclude the presence of a BSI and does not support withdrawal of antibiotics in clinically suspected septicemia. Even a sensitive PCR-based assay system may occasionally fail to detect the allegedly responsible microorganism in severely septic patients.20,22,25 Another reason one should be cautious in interpreting negative multiplex PCR results is that the SeptiFast detection list does not cover all BSI-causing pathogens. This, besides emphasizing the need for a combined use of SeptiFast and blood culture, also indicates that diagnosis of infection should not rely only on PCR results but should always incorporate any additional clinical and laboratory data available. One limitation of our study is that it was a nonrandomized, single-center study with a small number of cases, which may restrict generalization of results. Despite the small numbers, selection of cases investigated using the molecular methodology was based on clinical needs, and therefore our results may better reflect clinical reality as compared with random use of the test for pre-defined subjects. In conclusion, testing thoracic organ transplant recipients suspected for BSI for the presence of microbes by multiplex PCR improves pathogen detection and may help in
Chaidaroglou et al.
Multiplex PCR for Pathogen Detection
713
optimizing anti-microbial treatment of these severely immunocompromised patients. Use of the molecular methodology is suggested, especially in the early post-transplant period when recipients are at higher risk of infection and are in their most vulnerable post-surgical phase. Its capacity to identify blood-borne pathogens within 6 hours combined with its greater sensitivity justify the use of the multiplex PCR method for earlier diagnosis and prompt treatment of infection with appropriate antibiotics, which could lead to better clinical outcomes. However, because SeptiFast cannot detect a number of pathogens or determine antibiotic susceptibility, it should always be used in conjunction with traditional blood culture methods.
9. Garnacho-Montero J, Ortiz-Leyba C, Herrera-Melero I, et al. Mortality and morbidity attributable to inadequate empirical antimicrobial therapy in patients admitted to the ICU with sepsis: a matched cohort study. J Antimicrob Chemother 2008;61:436-41. 10. Peters RP, van Agtmael MA, Danner SA, et al. New developments in the diagnosis of bloodstream infections. Lancet Infect Dis 2004;4: 751-60. 11. Croxatto A, Prod’hom G, Greub G. Applications of MALDI-TOF mass spectrometry in clinical diagnostic microbiology. FEMS Microbiol Rev 2012;36:380-407. 12. Lehmann LE, Hunfeld KP, Emrich T, et al. A multiplex real-time PCR assay for the rapid detection and differentiation of 25 bacterial and fungal pathogens from whole blood samples. Med Microbiol Immunol 2008;197:313-24. 13. Lehmann LE, Alvarez J, Hunfeld KP, et al. Potential clinical utility of polymerase chain reaction in microbiological testing for sepsis. Crit Care Med 2009;37:3085-90. 14. American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med, 1992;20:864-74. 15. Garner JS, Jarvis WR, Emori TG, et al. CDC definitions for nosocomial infections. Am J Infect Control 1988;16:128-40. 16. National Nosocomial Infections Surveillance System. National Nosocomial Infections Surveillance System report: data summary from January 1992 to June 2002. Am J Infect Control 2002;30:458-75. 17. Hall KK, Lyman JA. Updated review of blood culture contamination. Clin Microbiol Rev 2006;19:788-802. 18. Snydman DR. Epidemiology of infections after solid-organ transplantation. Clin Infect Dis 2001;33 Suppl 1:S5-8. 19. Haddad F, Deuse T, Pham M, et al. Changing trends in infectious disease in heart transplantation. J Heart Lung Transplant 2010;29:306-15. 20. Louie RF, Tang Z, Albertson TE, et al. Multiplex polymerase chain reaction detection enhancement of bacteremia and fungemia. Crit Care Med 2008;36:1487-92. 21. Lehmann LE, Hunfeld KP, Steinbrucker M, et al. Improved detection of blood stream pathogens by real-time PCR in severe sepsis. Intensive Care Med 2010;36:49-56. 22. Bloos F, Hinder F, Becker K, et al. A multicenter trial to compare blood culture with polymerase chain reaction in severe human sepsis. Intensive Care Med 2010;36:241-7. 23. Bravo D, Blanquer J, Tormo M, 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. 24. Marik PE. Fungal infections in solid organ transplantation. Expert Opin Pharmacother 2006;7:297-305. 25. Tsalik EL, Jones D, Nicholson B, et al. Multiplex PCR to diagnose bloodstream infections in patients admitted from the emergency department with sepsis. J Clin Microbiol 2010;48:26-33.
Disclosure statement The authors have no conflicts of interest to disclose. The work was presented in part at the 31st annual meeting and scientific sessions of the International Society for Heart and Lung Transplantation, April 2011, San Diego, California.
References 1. Mattner F, Fischer S, Weissbrodt H, et al. Post-operative nosocomial infections after lung and heart transplantation. J Heart Lung Transplant 2007;26:241-9. 2. Kramer MR, Marshall SE, Starnes VA, et al. Infectious complications in heart–lung transplant. Analysis of 200 episodes. Arch Intern Med 1993;153:2010-6. 3. Palmer SM, Alexander BD, Sanders LL, et al. Significance of bloodstream infection after lung transplantation: analysis of 176 consecutive patients. Transplantation 2000;69:2360-6. 4. van de Beek D, Kremers WK, Del Roso JL, et al. Effect of infectious diseases on outcome after heart transplantation. Mayo Clin Proc 2008;83:304-8. 5. Rodriguez C, Muñoz P, Rodriguez-Créixems M, et al. Bloodstream infections among heart transplant recipients. Transplantation 2006;81:384-91. 6. Ibrahim EH, Sherman G, Ward S, et al. The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest 2000;118:146-55. 7. MacArthur RD, Miller M, Albertson T, et al. Adequacy of early empiric antibiotic treatment and survival in severe sepsis: experience from the MONARCS trial. Clin Infect Dis 2004;38:284-8. 8. Kollef MH, Sherman G, Ward S, et al. Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest 1999;115:462-74.