|Home | About | Journals | Submit | Contact Us | Français|
Prior to initiating a phase 1 dose escalation trial of the safety and immunogenicity of live, oral, recombinant, attenuated Salmonella enterica serovar Typhi vaccine strains in human subjects, the suitability of conventional blood culture procedures to rapidly and reliably detect the organisms in human blood was investigated. Blood culture specimens, with and without added growth supplements, were inoculated with study organism concentrations ranging from approximately 300 to as few as 1 to 2 CFU/10 ml culture and processed in a Bactec 9240 fluorescent series aerobic blood culture system. All cultures seeded with >6 CFU and 93% of cultures seeded with ~1 to 2 CFU were identified as positive for microbial growth within 44 h of incubation. The results were within the performance standard of ≤5 days to detection that is expected for Gram-negative cultures seeded at 10 to 50 CFU/vial. Recovery of test organisms from blood culture was not improved by the addition of supplements, but cultures with added supplements were identified positive an average of 5 h sooner than those without added supplements. Reliable detection of the investigational vaccine strains at <1 CFU/ml of blood within 2 days in conventional blood culture without added supplements allowed for shortened confinement time of study volunteers without compromising subject safety.
The ability to sensitively and rapidly monitor for bacteremia caused by live oral vaccine organisms is crucial for the evaluation of the safety of investigational vaccines in human subjects during clinical trials. Prior to initiating a phase 1 dose escalation trial of the safety and immunogenicity of live, oral, recombinant, attenuated Salmonella enterica serovar Typhi (S. Typhi) vector strains in human subjects, the suitability of conventional blood culture procedures to identify bacteremia caused by these investigational organisms was investigated.
Blood culture is generally considered the standard method for diagnosis of typhoid bacteremia but has been shown to detect only 40 to 70% of typhoid patients (1). In a quantitative study of bacteremia in typhoid fever patients, 50% of patients identified by positive blood culture had less than 1 CFU/ml of blood (2), illustrating the need for highly sensitive methods of detection. Rubin et al. (1) described methods to improve detection of S. Typhi in blood by use of a radiolabeled DNA probe specific for the genes encoding the Vi antigen present in concentrated blood samples. These methods were shown to detect <1 CFU/ml of S. Typhi in bacteremic typhoid patients; however, these methods are not currently applied in most clinical laboratories.
Advances in blood culture technology in the clinical lab have greatly enhanced the speed and sensitivity of detection of a broad range of clinically significant pathogens in blood (3). The current standard, Bactec Standard/10 Aerobic/F culture vials, together with Bactec 9240 brand fluorescent series instruments (Becton, Dickinson Diagnostic Instrument Systems, Sparks, MD), provide a sensitive, continuous monitoring blood culture system for aerobic culture and recovery of bacteria, yeast, and fungi from blood (4) and are routinely used in 60 to 70% of clinical labs (Becton, Dickinson Diagnostic Instruments Systems, Sparks, MD, personal communication). For ATCC reference and wild-type Gram-negative organisms seeded at 10 to 50 CFU/vial, this blood culture system has an expected time to detection of ≤5 days (5). Comparative clinical studies have shown that an incubation period of more than 5 days does not significantly improve recovery (4, 6). Therefore, it is standard clinical practice that blood cultures be monitored for 5 full days before a report of negative is generated.
Our phase 1 study protocol specified that volunteers be confined to an inpatient clinical facility for observation and monitoring after receipt of the investigational vaccines. Blood cultures would be collected twice daily for 7 days after vaccination to monitor for potential typhoid bacteremia. Two negative blood cultures on the 7th day postvaccination were required for release of subjects from confinement. Since clinical blood cultures are monitored for 5 full days before they are reported as negative, study volunteers could be confined in a clinical research facility for up to 2 weeks after vaccination pending issue of final blood culture results.
The ability of routine clinical blood culture to detect the investigational S. Typhi organisms was unknown. The vaccine strains used in this study are attenuated, have nutritional limitations, and have been shown to readily succumb to the bactericidal components in fresh human blood (7). Preliminary observations of enhanced survival after 1 h of incubation in blood culture with added supplements, targeted to support growth and fitness of the attenuated strains, compared to nonsupplemented blood culture (data not shown), suggested a possible benefit of additional supplements to blood culture vials for survival and growth of these investigational strains.
The objectives of this laboratory study were 2-fold. The first objective was to evaluate the survival, time to positive culture (TTP), and limits of detection in human blood of three investigational, recombinant, attenuated S. Typhi vector strains expressing a gene for Streptococcus pneumoniae surface protein antigen PspA (RASV-Sp) (7, 8, 9). Testing was done using the standard Bactec 9240 fluorescent series aerobic blood culture system. The second objective was to determine if recovery or time to detection could be improved by the addition of supplements to support rapid growth of the test organisms in cultures seeded with the lowest possible inoculum (~1 to 2 CFU/vial). These determinations would provide guidance as to the ability of standard clinical blood culture procedures to reliably and rapidly identify bacteremia caused by the attenuated investigational organisms and if supplements were needed to achieve or enhance this capability.
Three RASV-Sp strains containing plasmid pYA4088, specifying synthesis of the PspA antigen, were evaluated. These strains differ in parent strain source and RpoS status and have various requirements for optimal growth in culture (7, 8, 9).
χ9633(pYA4088) (S. Typhi ISP1820) is a live, recombinant, attenuated strain with genotype ΔPcrp527::TT araC PBAD crp ΔPfur81::TT araC PBAD fur Δpmi-2426 Δ(gmd-fcl)-26 ΔsopB1925 ΔrelA198::araC PBAD lacI TT ΔaraE25 ΔaraBAD23 ΔtviABCDE10 ΔagfBAC811 ΔasdA33.
χ9639(pYA4088) (S. Typhi Ty2 RpoS−) is a live, recombinant, attenuated strain with the genotype ΔPcrp527::TT araC PBAD crp ΔPfur81::TT araC PBAD fur Δpmi-2426 Δ(gmd-fcl)-26 ΔsopB1925 ΔrelA198::araC PBAD lacI TT ΔaraE25 ΔtviABCDE10 ΔagfBAC811 ΔasdA33 RpoS−.
χ9640(pYA4088) (S. Typhi Ty2 RpoS+) is a live, recombinant, attenuated strain with the genotype ΔPcrp527::TT araC PBAD crp ΔPfur81::TT araC PBAD fur Δpmi-2426 Δ(gmd-fcl)-26 ΔsopB1925 ΔrelA198::araC PBAD lacI TT ΔaraE25 ΔtviABCDE10 ΔagfBAC811 ΔasdA33 RpoS+.
Overnight cultures of χ9633(pYA4088), χ9639(pYA4088), and χ9640(pYA4088) were prepared from frozen vaccine seed stocks and incubated at 37°C with aeration in a phosphate-buffered, vegetable-based proprietary medium (KT broth). The growth medium for all RASV-Sp strains was supplemented with 0.1% d-(+)-mannose, 0.05% l-(+)-arabinose, 20 μg/ml l-tryptophan, and 22 μg/ml l-cysteine HCl. Strains χ9639(pYA4088) and χ9640(pYA4088) also received 0.1% d-(+)-glucose in all growth media (supplements were obtained from Sigma, St. Louis, MO, USA). The following day, overnight cultures were subcultured to fresh medium, supplemented as described above, and incubated for approximately 4 h at 37°C with aeration of 200 rpm until late-log-phase growth (an optical density at 600 nm of ~2.0, which corresponds to a culture density of 1 × 109 CFU/ml). Serial dilutions of each culture were prepared in phosphate-buffered saline (PBS) to generate inocula for blood culture vials. Ten-fold dilutions were prepared by adding 0.5 ml of harvested culture to 4.5 ml of PBS in sterile-capped 16- by 150-mm glass tubes, mixing, and transferring in series to achieve an estimated culture density of 1,000 and 100 CFU/ml in the 6th and 7th tubes, respectively. Further downstream dilutions were made in larger volumes to prepare CFU/ml inocula ranging from approximately 200 to ~1 to 2 CFU/ml for seeding replicates of 10-ml test blood culture specimens, representing a bacterial concentration range of approximately 20 to 0.1 CFU/ml of blood collected in a clinical test sample.
Bactec Standard/10 Aerobic/F culture vials (here referred to as Bactec vials or vials) containing 40 ml F medium comprised of enriched soybean-casein digest broth dispensed with added CO2 were used for all experiments. Five normal, healthy volunteers (3 male and 2 female; age, 20 to 58) gave informed consent and provided blood for the experiments, in compliance with protocol no. 12450, approved by the Saint Louis University (SLU) Institutional Review Board. Approximately 10 ml of blood was collected directly into each Bactec test vial. The vials were transferred immediately to the research lab, where each was labeled to identify test condition, without indicating RASV-Sp strain, dose of inoculum, or the presence or absence of supplements. A supplement mixture comprised of d-(+)-mannose, l-(+)-arabinose, l-tryptophan, and l-cysteine HCl was added to vials of the “+ supplements” group to provide a final concentration in culture of 0.1%, 0.05%, 20 μg/ml, and 22 μg/ml, respectively. Freshly prepared RASV-Sp suspensions of titrated CFU/ml were drawn up in sterile syringes and inoculated into the designated vials. Blood culture preparations were kept at ambient temperature (15 to 30°C) until collected by a laboratory courier within 2 h of inoculation. The cultures were packaged in an insulated pouch and transported by car at ambient temperature to Quest Diagnostics, St. Louis, MO (a distance of 17 miles; approximately 25 min of driving time). The time from inoculation of blood cultures to arrival at the testing laboratory was approximately 2.5 h. On receipt at Quest Diagnostics, blood culture vials were inserted into a Bactec 9240 fluorescent series instrument, where they were incubated, together with normal laboratory controls, for up to 5 full days or until generating a positive signal (whichever was sooner). The instrument was a fluorescence detector which incubates (35 ± 1°C) and rocks the culture vials while continuously monitoring for fluorescence, which was proportional to the amount of CO2 in the culture. Each individual vial was monitored every 10 min. Increases in CO2 generated a signal indicating the presence of microorganisms metabolizing media in the blood culture, identifying a sample as presumptively positive. Time from insertion of the vials in the instrument to the detection of a positive signal for each culture was documented. Cultures signaling CO2 increase were removed from the instrument and subcultured to blood agar for confirmation of positive culture and Gram stain testing. All plated cultures recovered from Bactec blood culture vials were retrieved by SLU laboratory personnel to evaluate for RASV-Sp strain colony morphology and typing by agglutination with Salmonella O, factor 9, group D antiserum (Difco, Detroit, MI, USA).
Primary culture media, subculture media, and PBS used for RASV-Sp growth, suspension, and titration were spread onto tryptic soy agar (TSA; Difco, Detroit, MI, USA) plates to monitor for contamination. Dilutions of RASV-Sp strain preparations were spread on TSA for quantification of the approximate number of CFU that were inoculated into each test vial. Samples of each RASV-Sp strain preparation were streaked for CFU isolation onto TSA, MacConkey agar (Difco, Detroit, MI, USA) supplemented with 1% maltose (MMA), and MacConkey agar supplemented with 1% maltose and 1% arabinose (MAMA). After overnight incubation at 37°C, media and PBS control TSA plates were observed for extraneous organisms. Plates of titrated RASV-Sp inocula were observed for single-culture morphology, RASV-Sp strain phenotype, and CFU count. Counts from TSA plates yielding 30 to 300 CFU were used to determine approximate CFU added to each test vial. TSA, MMA, and MAMA plates streaked for isolation of CFU were observed for the typical phenotype of each RASV-Sp strain, and selected colonies from test preparations and isolates recovered from blood culture were tested by Gram stain and agglutination with Salmonella O, factor 9, group D antiserum (Difco, Detroit, MI, USA) to confirm Gram-negative morphology and positive agglutination with type-specific antiserum.
Arithmetic means were used to calculate averages. The chi-square test was used to compare recovery of cultures between groups. The Mann-Whitney U test, used to compare time to detection between groups, was performed using GraphPad Prism software version 5.03 for windows (GraphPad Software, San Diego, CA, USA). The Poisson distribution analysis was used to predict the number of vials containing the test organisms.
Experiments were designed to test the RASV-Sp strains at a range of concentrations in blood culture. In the first experiments, 24 blood culture vials (eight for each of 3 RASV-Sp strains) were prepared with fresh blood from a single volunteer. Suspensions of each RASV-Sp subculture were prepared as described above to approximate 200 CFU/ml, and serial 1:3 dilutions were prepared in PBS to generate suspensions of approximately 66 CFU/ml, 22 CFU/ml, and 7 CFU/ml.
Because the RASV-Sp strains are tryptophan and cysteine auxotrophs and possess a ΔPcrp527::TT araC PBAD crp mutation which imposes a growth defect in the absence of arabinose, we tested to see if the addition of these supplements to the culture medium would improve recovery of the RASV-Sp strains. Mannose was supplied to allow for complete lipopolysaccharide (LPS) O-antigen synthesis during blood culture.
For each RASV-Sp strain, 4 vials were prepared with added supplements and 4 without added supplements. One milliliter of each seed suspension was inoculated into 1 supplemented and nonsupplemented vial before transfer to Quest Diagnostics for processing.
After confirmation that test cultures could be detected at higher concentrations, experiments were performed to test the RASV-Sp strains in blood culture at the lowest possible inoculum of 1 CFU/vial. Twenty blood culture vials for each of three RASV-Sp strains were prepared (one strain per day on three consecutive days), using fresh blood from a different individual each day. For each RASV-Sp series, 10 vials were prepared with added supplements and 10 without added supplements. Suspensions of RASV-Sp strains were prepared as described above to approximately 10 CFU/ml and 1 CFU/ml. Three of each supplemented and nonsupplemented vials for each strain were inoculated with a 1-ml suspension containing ~10 CFU as low positive controls. Seven of each supplemented and nonsupplemented vials for each strain were inoculated with 1 ml suspension containing ~1 CFU. Culture vials were transferred to Quest Diagnostics for processing as described above. As an internal lab control for the presence of bacteria in ~1 CFU/ml seed suspensions, 1 ml of each RASV-Sp strain suspension containing ~1 CFU/ml was added to 3 tubes each of 40 ml phosphate-buffered KT broth at the same time the Bactec vials were inoculated. These permissive control cultures were retained and incubated at 37°C in the SLU research lab and observed for visible growth after incubation for 36 to 72 h.
Quality assurance negative-control plates for all experiments showed no evidence of contaminating organisms on plates spread with primary medium, subculture medium, or PBS used in any RASV-Sp strain culture and inoculum preparations. All seed preparations of RASV-Sp strains spread to TSA, MMA, and MAMA showed the typical strain phenotype, were Gram negative, and were positive for agglutination with Salmonella O, factor 9, group D antiserum, confirming RASV-Sp strain identities. All blood cultures giving a positive signal and subcultured to blood agar at Quest Laboratory were recovered and confirmed as pure Gram-negative cultures with the typical phenotype for the RASV-Sp strains. A subset of 9 cultures (three of each RASV-Sp strain) was further tested for RASV-Sp strain identity and confirmed positive for agglutination with Salmonella O, factor 9, group D antiserum. Internal lab controls of ~1 CFU/ml RASV-Sp seed suspensions when inoculated into permissive KT broth media showed growth by visual cloudiness in 8 of 9 cultures within 72 h of incubation at 37°C.
Counts of titrated and plated RASV-Sp seed preparations showed that higher-dose test vials were inoculated with average titrated doses of 272, 91, 30, and 10 CFU/vial, representing concentrations in blood of 27, 9, 3, and 1 CFU/ml. Low-dose vials were seeded with 2.3, 1.6, and 1.2 CFU/vial for χ9633(pYA4088), χ9639(pYA4088), and χ9640(pYA4088), respectively, achieving the minimum possible seed doses of ~1 to 2 CFU/vial, equivalent to ~0.1 to 0.2 CFU/ml of blood for all three strains.
A rapid and high rate of recovery of viable test organisms from blood cultures seeded with any of the RASV-Sp strains was observed, with and without added supplements. Eighty-one of 84 blood culture vials tested in combined experiments seeded with any RASV-Sp strain in doses ranging from 1.2 to 355 CFU/10 ml of blood were detected as positive for live organisms within 44 h of culture, giving an overall 96% recovery rate in less than 2 days of incubation. All 42 vials seeded with ≥6 CFU/vial were detected within 39 h of incubation. At the lowest inoculum of ~1 to 2 CFU/vial, 39 of 42 cultures (93%) were recovered and were all detected within 44 h of incubation. Since the seed suspensions used to inoculate the lowest inoculum blood cultures contained only ~1 to 2 CFU/ml, it is possible that no organisms were present in the 1 ml inoculum used to seed the 3 vials that remained negative after 5 days incubation, and these results may have been true negatives. Using Poisson distribution analysis, one would expect two to three negative cultures.
Individual TTP results for blood cultures tested under standard clinical conditions (without added supplements) are shown by RASV-Sp strain and by calculated seed dosages in Table 1 (for vials seeded in titration from 355 to 6 CFU) and Table 2 (for vials seeded with ~1 to 2 CFU).
Recovery was not significantly enhanced by addition of nutritional supplements for any of the RASV-Sp strains, with overall recovery of 40/42 cultures without added supplements, compared to 41/42 with added supplements (P > 0.5).
For samples inoculated with ~1 to 2 CFU/vial, 19/21 vials without added supplements were recovered [6/7 χ9633(pYA4088) and χ9639(pYA4088) and 7/7 χ9640(pYA4088)] compared to 20/21 vials with added supplements [6/7 χ9639(pYA4088) and 7/7 χ9633(pYA4088) and χ9640(pYA4088)] (Fig. 1). These differences were not statistically significant (P > 0.5).
While the addition of supplements to the blood culture media was not necessary for recovery, added supplements had a small but significant effect on the TTP of all three RASV-Sp strains in blood culture. A comparison of results between nonsupplemented and supplemented blood cultures by the RASV-Sp strain, when seeded with ~1 to 2 CFU/vial, is shown in Fig. 1. The average TTP for supplemented blood cultures inoculated with ~1 to 2 CFU/vial was 29 h (range, 23.4 to 35.9) versus 34.5 h (range, 27.3 to 43.6) for the cultures without added supplements. Supplemented cultures containing only ~1 to 2 CFU signaled positive an average 5.7 h sooner than nonsupplemented cultures: Δ8.1 h for χ9633(pYA4088) (P = 0.002), Δ4.9 h for χ9639(pYA4088) (P = 0.008), and Δ4.3 h for χ9640(pYA4088) (P = 0.018).
Results comparing TTP for all 81 recovered cultures with and without added supplements, and seeded with a range from 1.2 to 355 CFU/vial, are shown together in Fig. 2. The average time to detection for 41 cultures with added supplements was 27 h (range, 19.6 to 35.8), compared to an average of 32 h (range, 19.3 to 43.6) to detection for the 40 cultures without added supplements. Overall, an average reduction in TTP of 5 h was observed in vials with added supplements, which is a 16% reduction in TTP compared to that of vials without added supplements (P < 0.001).
The manufacturer's expected performance parameters using Standard/10 Aerobic/F culture vials are that various ATCC and wild-type organisms, seeded at 10 to 50 CFU/vial, can be detected within 5 days of incubation (5). For the studies reported here, all positive blood cultures, including the cultures seeded with ~1 to 2 CFU of RASV-Sp strains/10 ml blood, with or without added supplements, were identified within 2 days of incubation.
Recovery of pathogens in blood culture can be affected by multiple parameters in addition to bacterial concentration, including blood volume, culture medium composition and volume, and incubation time (10). Successful recovery of these attenuated RASV-Sp strains from blood culture without added supplements may be due in part to the enrichment factors present in F medium, which include yeast extract, animal tissue digest, sucrose, hemin, menadione, and vitamin B6, as well as nutritional components available in human blood. RASV-Sp strains are sensitive to inhibiting factors found in fresh blood, including complement components (7). However, the dilution of blood specimens in broth at a ratio of 1:5, as was used in these studies, may mitigate this effect and has been shown to be optimal to support growth of most organisms in blood culture (10). Clinical blood culture media and culture conditions are developed and optimized to recover a broad spectrum of possible disease-causing organisms, including slow-growing, fastidious species. Results reported here for our investigational strains are consistent with other published studies, indicating that incubation time of fewer than 5 days may be sufficient for recovery of many clinically significant organisms (11, 12, 13, 14).
These results demonstrate that conventional blood culture methods, using the Bactec 9240 fluorescent series blood culture system and Standard/10 Aerobic/F culture vials, provided reliable recovery of three recombinant attenuated S. Typhi vector vaccine organisms from human blood inoculated with ~1 to 2 CFU/10 ml blood (<1 CFU/ml), within 2 days of incubation, and without added supplements. This allows a more accurate evaluation of the clinical behavior of live, attenuated S. Typhi strains and also decreases the risk associated with administering such strains in a clinical trial setting. The addition of supplements provided a small but significant reduction in the time to detection but did not enhance recovery over nonsupplemented cultures. The potential benefit of supplementation in order to achieve more rapid detection of a test organism should be weighed against the risk of introducing contaminants when adding materials to sterile media in culture vials that have been manufactured under highly controlled conditions. Further studies should be undertaken to confirm and extend these findings in other trials and to validate this procedure for clinical use.
The results of these studies provided guidance for conduct of a phase 1 vaccine trial of live oral S. Typhi vaccine strains. Study volunteers were subjected to fewer days of confinement which, in turn, reduced study costs associated with extended confinements. The sensitivity and rapidity of detection enabled by the Bactec 9240 blood culture system provides an important and broadly available tool to support future studies of the safety of recombinant, attenuated S. Typhi vaccines in human volunteers.
This study was funded by The Bill and Melinda Gates Foundation Grand Challenges in Global Health grant number 37863.
We thank Lisa Michelson, at Quest Diagnostics, for technical assistance, the nurses and technologists for their efforts, and the volunteers who provided blood for these studies.
Published ahead of print 24 July 2013