The Fate of Bacteria Introduced into Whole Blood from Which Platelet Concentrates Were Prepared and Stored at 22 or 4 C R. A. KAHNAND R. L. SYRING From the Blood Research Loborotory. American National Red Cross, Bethesdo. Morylond
whole blood which were used for platelet concentrates. The platelets were stored either at room temperature (22 C) or at 4 C.
Bacteria were intentionally introduced into units of whole blood. Platelet concentrates (PC) which were made from these units were stored at either room temperature (22 C ) or at 4 C. In order to isolate small numbersof bacteria from a P C (i.e., 1 to 10 organisms per ml), substantial contamination (42 to 125 organisms per ml) of the whole blood was required. If the P C were stored at room temperature, all organisms except Pseudomonas aeruginosa, which was apparently killed, grew out of control within 48 hours. Storage of P C at 4 C resulted in the general maintenance of bacterial numbers. Since gross contamination of P C has only occasionally been reported, we conclude that past reports of modest contamination of ptatelet concentrates are primarily sampling artifacts.
Materials and Methods
Units of whole blood for the study were obtained from the District of Columbia chapter of the American National Red Cross. The blood was drawn from healthy volunteers into C P D anticoagulant. Immediately following collection, the blood was allowed to sit undisturbed at room temperature until the experiment began. This interval was usually two hours but no longer than A NUMBER of conflicting reports have been four hours. published regarding the possible bacterial The bacteria used in these experiments were all stock culture strains of Escherichia coli. Staphycontamination of units of platelet conlococcus epidermidis, Enterobacter cloacae, and centrates. Some investigators have reported Pseudomonas aeruginosa. These organisms were no c o n t a m i n a t i ~ n , ~ while ~ ' ~ ~ others ~ ~ ~ ~ ~ ~ chosen ~~ because they have been identified as have reported that from 1 to 6 per cent of contaminants of either whole blood1.13or platelet the platelet concentrates cultured were concentrate^.'.^ The stock cultures were maintained on trypticase soy agar (BBL) slants at 4 C found to contain b a c t e ~ i a . ~Much . ~ . ~ of the and transferred biweekly. The organism selected controversy centers around methods of for a particular experiment was grown a t 32 C for sampling and the possibility of accidental 24 hours on trypticase soy agar plates and washed contamination. Another possible source of from the surface of the plates with sterile saline contamination is the skin plug which can immediately before use. The bacteria were then diluted with sterile saline and 0.5 ml was removed enter the blood bag during phlebotomy.' It for use as the inoculum. All inoculations, subsewould seem inevitable that at least some quent sampling and plating were carried out phlebotomies will be initially contaminated aseptically and the procedures performed in a in this way. This raises questions regarding laminar air-flow hood. the fate of organisms contaminating the The number of bacteria per ml in whole blood or platelet concentrate was determined by diswhole blood from which platelet contributing a 3 ml sample equally on three plates of centrates are prepared. We have studied the trypticase soy agar. When platelet concentrates fate of bacteria introduced into units of were sampled, the aliquot removed from the unit was mixed well on a vortex stirrer before plating. Received for publication September 19, 1974; acAll plates were incubated at 32 C for 24 hours becepted December 10, 1974. fore being counted with a Quebec Colony Counter Contribution No. 295 from the American National (American Optical Corp.) The number of coloRed Cross Blood Research Laboratory. Supported in nies on the three plates were averaged to deterpart by a Washington Heart Association Research Fellowship awarded to R. A. Kahn. mine the average number of colony forming units
363 Transfusion July-Aug. 1975
Volume I5 Number4
364
Transfusion July-Aug. 1975
K A H N A N D SYRING
of bacteria per ml of whole blood or platelet concentrate. In some samples, only one colony was seen for all three plates, and the result was expressed as < l/ml. Units of whole blood were inoculated through a port by means of a medication injection site (Fenwal AE-7). The blood was then mixed well, samples were taken for plating, and the units were centrifuged. As a control with each experimental unit, 500 ml of sterile saline were inoculated with the same bacterial preparation and a sample taken for bacterial plating. Centrifugation to obtain platelet rich plasma was done a t 22 C in a Sorvall RC-3 centrifuge a t 2025 x g (2800 rpm) for one minute and 40 seconds a t speed. Plateletrich plasma was then extracted into a satellite bag and centrifuged a t 3164 x g (3500 rpm) for 10 minutes a t speed at 22 C. From 24 to 32 ml of plasma was left with the platelets. The concentrate was allowed to sit undisturbed a t room temperature for 45 minutes and then was gently resuspended. Those units which showed macroscopic platelet clumping were discarded. Following resuspension, samples were taken (via a medication injection site) for bacterial studies. The unit was then stored for three days either at room temperature with continuous agitation or at 4 C undisturbed. Samples were taken for plating every 24 hours. No evidence of accidental contamination was found in the course of sampling or plating and none of the units showed any evidence of bacterial contamination with organisms other than those purposely introduced.
Results In order to obtain small numbers of bacteria from the platelet concentrates, it was necessary to inoculate the whole blood with large numbers of bacteria (Table I). The possibility that bacteria were killed in the process of centrifugation and preparation of the concentrate was investigated by also culturing the red blood cells and platelet poor plasma obtained from a number of units. It was found that the sum of the concentration of bacteria found in each component was nearly identical to the concentration of bacteria in the whole blood. T o rule out a bactericidal effect of the whole blood from the time of inoculation until sampling, 500 ml of saline was simultaneously inoculated and cultured along with the whole blood. From this experiment it was determined that, in the limited time interval that the bacteria were exposed to the whole blood, no bactericidal effect occurred. Room Temperature (22 C )Storage With two of the four organisms used (Escherichia coli and Enterobacter cloacae) considerable
Table 1. Fate of Bacteria Introduced into Whole Blood (WB) from which Platelet Concentrates (PC) Were Made
Organism
WE
PC
(Avg. No. of Eacterialml and Range)
(Avg. No. of Eacterialml and Range)
Escherichia coli 125 (78-161) Staphylococcus 42 (9-79) epidermidis 94 (39-152) Enterobacter cloacae Pseudomonasaeruginosa 100 (47-128)
3 (.3-11) < 1 (0-3) 11 (2-40) 24 (0-104)
'Values represent the average number of cotony forming units of bacteria inoculated into whole blood and subsequently found in platelet concentrates. Platelet concentrates were sampled immediately following resuspension. A t least nine experiments were conducted with each organism.
growth was seen after 24 hours of storage (Table 2). However, with Staphylococcus epidermidis generally minimal growth took place after 24 hours, but substantial growth took place after 48 hours of storage. When Pseudomonas aeruginosa was used, there was evidence of a bactericidal effect. In most of these platelet concentrates, no organisms could be isolated after 24 hours and all units were apparently sterile after 48 hours of storage. Refrigerated ( 4 C )Storage A bacteriostatic effect was seen in concentrates contaminated with either Escherichia coli or Staphylococcus epidermidis and stored for three days a t 4 C (Table 3). When Enterobacter cloacae or Pseudomonas aeruginosa were the contaminating organisms, the number of viable bacteria per ml of concentrate decreased over the three-day storage interval, especially when the latter organism was used. However, even a t the end of the storage period all units were still contaminated with bacteria.
Discussion
In order to isolate small numbers (1 to 10/ml) of bacteria from a platelet concentrate immediately following preparation, substantial numbers (42 to 125/ml) of bacteria must have been present in the whole blood. This could be explained on the basis of sedimentation properties of bacteria which may be different from those of platelets or the masking of the presence of bacteria by platelets via platelet aggregation, a
Volume 15 Number4
365
BACTERIA IN PLATELET CONCENTRATES Table 2. Number o f Colony Forming Units of Bacteria Per ml of Platelet Concentrate (PC) Stored at Room Temperature (22CI ’
0raa nism Escherichia coli (N = 5)
PC = 0 Hour
Day 1
Day 2
Day 3
3 3
2 Staphylococcus epidermidis (N = 6 )
>300 >300 31 15 0 1 2 0
2 300 >300 59
>300
>300
4 Pseudomonas aeruginosa (N = 5)
8 0 0
>300 0 300 >300 >300
>300 >300
>300
>300
>300
>300
Enterobacter cloacae (N = 6)
>300)
’
1} J
>300
0
0 0
0 0
0 0
0
0
0
‘All values reflect the average number of bacteria from triplicate 1 ml samples. Samples were taken immediately following resuspension of the platelet concentrate (PC = 0 hour) and every 24 hours for three days (days 1-3). Values of >300 represent overgrowth on all plates.
phenomenon which has been d ~ c u m e n t e d . ~ We suspect the former mechanism is predominant since bacteria could be isolated from both red blood cells and platelet-poor plasma and when the bacterial concentration of all three components are added the value is nearly identical to the concentration found
in the whole blood. Regardless of the mechanism, the fact remains that few bacteria could be found in the platelet concentrates compared to the concentration found in the initial whole blood. Investigators who have cultured platelet concentrates routinely and have found some to be
Table 3. Number of Colony Forming Units of Bacteria Per ml of Platelet Concentrate (PC) Stored at 4 C Organism Escherichia coli (N = 4)
Staphylococcus epidermidis (N = 41
PC = 0 Hour
Day 1
whole blood which were used for platelet concentrates. The platelets were stored either at room temperature (22 C) or at 4 C.
Bacteria were intentionally introduced into units of whole blood. Platelet concentrates (PC) which were made from these units were stored at either room temperature (22 C ) or at 4 C. In order to isolate small numbersof bacteria from a P C (i.e., 1 to 10 organisms per ml), substantial contamination (42 to 125 organisms per ml) of the whole blood was required. If the P C were stored at room temperature, all organisms except Pseudomonas aeruginosa, which was apparently killed, grew out of control within 48 hours. Storage of P C at 4 C resulted in the general maintenance of bacterial numbers. Since gross contamination of P C has only occasionally been reported, we conclude that past reports of modest contamination of ptatelet concentrates are primarily sampling artifacts.
Materials and Methods
Units of whole blood for the study were obtained from the District of Columbia chapter of the American National Red Cross. The blood was drawn from healthy volunteers into C P D anticoagulant. Immediately following collection, the blood was allowed to sit undisturbed at room temperature until the experiment began. This interval was usually two hours but no longer than A NUMBER of conflicting reports have been four hours. published regarding the possible bacterial The bacteria used in these experiments were all stock culture strains of Escherichia coli. Staphycontamination of units of platelet conlococcus epidermidis, Enterobacter cloacae, and centrates. Some investigators have reported Pseudomonas aeruginosa. These organisms were no c o n t a m i n a t i ~ n , ~ while ~ ' ~ ~ others ~ ~ ~ ~ ~ ~ chosen ~~ because they have been identified as have reported that from 1 to 6 per cent of contaminants of either whole blood1.13or platelet the platelet concentrates cultured were concentrate^.'.^ The stock cultures were maintained on trypticase soy agar (BBL) slants at 4 C found to contain b a c t e ~ i a . ~Much . ~ . ~ of the and transferred biweekly. The organism selected controversy centers around methods of for a particular experiment was grown a t 32 C for sampling and the possibility of accidental 24 hours on trypticase soy agar plates and washed contamination. Another possible source of from the surface of the plates with sterile saline contamination is the skin plug which can immediately before use. The bacteria were then diluted with sterile saline and 0.5 ml was removed enter the blood bag during phlebotomy.' It for use as the inoculum. All inoculations, subsewould seem inevitable that at least some quent sampling and plating were carried out phlebotomies will be initially contaminated aseptically and the procedures performed in a in this way. This raises questions regarding laminar air-flow hood. the fate of organisms contaminating the The number of bacteria per ml in whole blood or platelet concentrate was determined by diswhole blood from which platelet contributing a 3 ml sample equally on three plates of centrates are prepared. We have studied the trypticase soy agar. When platelet concentrates fate of bacteria introduced into units of were sampled, the aliquot removed from the unit was mixed well on a vortex stirrer before plating. Received for publication September 19, 1974; acAll plates were incubated at 32 C for 24 hours becepted December 10, 1974. fore being counted with a Quebec Colony Counter Contribution No. 295 from the American National (American Optical Corp.) The number of coloRed Cross Blood Research Laboratory. Supported in nies on the three plates were averaged to deterpart by a Washington Heart Association Research Fellowship awarded to R. A. Kahn. mine the average number of colony forming units
363 Transfusion July-Aug. 1975
Volume I5 Number4
364
Transfusion July-Aug. 1975
K A H N A N D SYRING
of bacteria per ml of whole blood or platelet concentrate. In some samples, only one colony was seen for all three plates, and the result was expressed as < l/ml. Units of whole blood were inoculated through a port by means of a medication injection site (Fenwal AE-7). The blood was then mixed well, samples were taken for plating, and the units were centrifuged. As a control with each experimental unit, 500 ml of sterile saline were inoculated with the same bacterial preparation and a sample taken for bacterial plating. Centrifugation to obtain platelet rich plasma was done a t 22 C in a Sorvall RC-3 centrifuge a t 2025 x g (2800 rpm) for one minute and 40 seconds a t speed. Plateletrich plasma was then extracted into a satellite bag and centrifuged a t 3164 x g (3500 rpm) for 10 minutes a t speed at 22 C. From 24 to 32 ml of plasma was left with the platelets. The concentrate was allowed to sit undisturbed a t room temperature for 45 minutes and then was gently resuspended. Those units which showed macroscopic platelet clumping were discarded. Following resuspension, samples were taken (via a medication injection site) for bacterial studies. The unit was then stored for three days either at room temperature with continuous agitation or at 4 C undisturbed. Samples were taken for plating every 24 hours. No evidence of accidental contamination was found in the course of sampling or plating and none of the units showed any evidence of bacterial contamination with organisms other than those purposely introduced.
Results In order to obtain small numbers of bacteria from the platelet concentrates, it was necessary to inoculate the whole blood with large numbers of bacteria (Table I). The possibility that bacteria were killed in the process of centrifugation and preparation of the concentrate was investigated by also culturing the red blood cells and platelet poor plasma obtained from a number of units. It was found that the sum of the concentration of bacteria found in each component was nearly identical to the concentration of bacteria in the whole blood. T o rule out a bactericidal effect of the whole blood from the time of inoculation until sampling, 500 ml of saline was simultaneously inoculated and cultured along with the whole blood. From this experiment it was determined that, in the limited time interval that the bacteria were exposed to the whole blood, no bactericidal effect occurred. Room Temperature (22 C )Storage With two of the four organisms used (Escherichia coli and Enterobacter cloacae) considerable
Table 1. Fate of Bacteria Introduced into Whole Blood (WB) from which Platelet Concentrates (PC) Were Made
Organism
WE
PC
(Avg. No. of Eacterialml and Range)
(Avg. No. of Eacterialml and Range)
Escherichia coli 125 (78-161) Staphylococcus 42 (9-79) epidermidis 94 (39-152) Enterobacter cloacae Pseudomonasaeruginosa 100 (47-128)
3 (.3-11) < 1 (0-3) 11 (2-40) 24 (0-104)
'Values represent the average number of cotony forming units of bacteria inoculated into whole blood and subsequently found in platelet concentrates. Platelet concentrates were sampled immediately following resuspension. A t least nine experiments were conducted with each organism.
growth was seen after 24 hours of storage (Table 2). However, with Staphylococcus epidermidis generally minimal growth took place after 24 hours, but substantial growth took place after 48 hours of storage. When Pseudomonas aeruginosa was used, there was evidence of a bactericidal effect. In most of these platelet concentrates, no organisms could be isolated after 24 hours and all units were apparently sterile after 48 hours of storage. Refrigerated ( 4 C )Storage A bacteriostatic effect was seen in concentrates contaminated with either Escherichia coli or Staphylococcus epidermidis and stored for three days a t 4 C (Table 3). When Enterobacter cloacae or Pseudomonas aeruginosa were the contaminating organisms, the number of viable bacteria per ml of concentrate decreased over the three-day storage interval, especially when the latter organism was used. However, even a t the end of the storage period all units were still contaminated with bacteria.
Discussion
In order to isolate small numbers (1 to 10/ml) of bacteria from a platelet concentrate immediately following preparation, substantial numbers (42 to 125/ml) of bacteria must have been present in the whole blood. This could be explained on the basis of sedimentation properties of bacteria which may be different from those of platelets or the masking of the presence of bacteria by platelets via platelet aggregation, a
Volume 15 Number4
365
BACTERIA IN PLATELET CONCENTRATES Table 2. Number o f Colony Forming Units of Bacteria Per ml of Platelet Concentrate (PC) Stored at Room Temperature (22CI ’
0raa nism Escherichia coli (N = 5)
PC = 0 Hour
Day 1
Day 2
Day 3
3 3
2 Staphylococcus epidermidis (N = 6 )
>300 >300 31 15 0 1 2 0
2 300 >300 59
>300
>300
4 Pseudomonas aeruginosa (N = 5)
8 0 0
>300 0 300 >300 >300
>300 >300
>300
>300
>300
>300
Enterobacter cloacae (N = 6)
>300)
’
1} J
>300
0
0 0
0 0
0 0
0
0
0
‘All values reflect the average number of bacteria from triplicate 1 ml samples. Samples were taken immediately following resuspension of the platelet concentrate (PC = 0 hour) and every 24 hours for three days (days 1-3). Values of >300 represent overgrowth on all plates.
phenomenon which has been d ~ c u m e n t e d . ~ We suspect the former mechanism is predominant since bacteria could be isolated from both red blood cells and platelet-poor plasma and when the bacterial concentration of all three components are added the value is nearly identical to the concentration found
in the whole blood. Regardless of the mechanism, the fact remains that few bacteria could be found in the platelet concentrates compared to the concentration found in the initial whole blood. Investigators who have cultured platelet concentrates routinely and have found some to be
Table 3. Number of Colony Forming Units of Bacteria Per ml of Platelet Concentrate (PC) Stored at 4 C Organism Escherichia coli (N = 4)
Staphylococcus epidermidis (N = 41
PC = 0 Hour
Day 1
