A blood banker in the Southwest USA reports that a pediatric patient received a transfusion of a pool of four platelet concentrates and experienced rigors and a rise in temperature to 40C. The four platelet concentrates were pooled in the morning at 0740 by a long term experienced technologist using a laminar flow bench, were then leukoreduced using an FDA approved filter, and finally were irradiated just before being issued for transfusion that same morning at 0900. The transfusion was stopped when the patient became symptomatic, and a Gram's stain of the residual pooled platelet product showed "numerous short Gram negative rods". The patient was started on antibiotics after blood cultures were drawn (the names of the antibiotics were not provided). The patient recovered within about 2 hours of the initial temperature spike and rigors. Cultures of the residual pooled platelet product and blood cultures of the patient were reported as showing "no growth" after 10 days. The above situation was reported to the blood collection agency. The platelet concentrates in the pool ranged in age from 3-5 days, so that the blood collection agency was able to have quarantined and cultured all of the untransfused RBC and plasma products associated with the four platelet concentrates in question. None of these cultures were positive, and the blood bank was told that it could not be confirmed that the reaction was due to bacterial contamination. The Southwestern blood banker wonders if the fact that the platelet pool was irradiated just before being issued for transfusion might have played a role in mitigating a septic transfusion reaction. What is the opinion of the e-network forum why numerous short Gram negative rods in the pooled platelet product did not cause a positive culture result? Assuming the usual dose of irradiation was delivered to the pooled platelets just before transfusion, what is the possibility that the irradiation step prevented the contaminating bacteria from growing in the culture? What other explanations might be plausible?

The following replies were submitted in response to the above case report and queries. The consensus opinion is that pre-transfusion irradiation of pooled platelets will NOT prevent growth of bacteria in culture, so that other explanations for lack of growth in culture must be considered, including a fastidious contaminating organism or even a false positive Gram stain result.

1. A medical epidemiologist at the Centers for Disease Control and Prevention, wrote that it is entirely possible for the reported reaction to have occurred solely due to endotoxin. The CDC epidemiologist adds that "Unfortunately, the presumed presence of endotoxin or of characterization of the gram-negative rods as "short" does not help much in narrowing down the differential. Short rods remind one microbiologist of Haemophilus spp., but I would think this quite unlikely in a blood product, but Enterobacteriacae can also be judged as quite short. As network readers are probably aware, E. coli, and Serratia and Enterobacter spp. are the most frequent gram-negative pathogens, although Yersinia enterocolitica has been reported in platelets. Of course, there may be other explanations for a negative culture - depending on the fastidiousness of the organism, and the conditions under which the product or the patient were cultured. If either the blood bag, patient blood culture, or patient serum (for endotoxin) were available, there are CDC laboratories that might be able to speciate on the basis of rRNA or other techniques."

2. A blood banker in Texas wrote that she has had experience working as a microbiologist as well as working as a blood banker. She points out that a Gram stain on a platelet concentrate can be problematic because platelet material and debris can be misinterpreted as bacteria. She commented that her degree of confidence in a 'positive' Gram stained slide result would be influenced accordingly if the slide was examined by a microbiology technologist who was extensively experienced in examining platelet concentrates. The Texan comments that this sort of scenario raises the kind of difficult questions that we will be confronting in the future, as we seek to deal with bacterial contamination. She adds that a blood culture is not like doing a simple blood type and that the answer is not necessarily straightforward. The culture may be positive depending on the level of bacterial growth at that time, the media used and the temperature at which the culture is run, the presence or absence of antibiotics or other bacteriostatic agents in the patient, anticoagulant, or media, etc. The presence of a positive culture may also be due to contamination occurring at the time of planting the culture.

3. A microbiologist from Connecticut wrote that the dose of irradiation used for inactivation of lymphocytes to prevent TA-GVHD is too low to kill a significant fraction of bacteria. Killing of living organisms by radiation depends on the 'target size' of the organism, which is essentially the size of the genome. A dose of radiation sufficient to produce a lethal dose of breaks in a 10E9 base human genome is insufficient to produce a lethal dose of breaks in the much smaller genome of the typical bacterium; the lethal radiation dose in bacteria is anywhere from 10-1000+-fold higher than for mammalian cells. This is an unlikely cause of negative cultures. As far as the lack of growth of the organisms seen on Gram stain, the microbiologist has several questions for the inquiring blood banker who provided the case to the e-network: Did several good microbiologists review the slide? Clinical microbiology labs are not used to examining platelet concentrates, for the most part. Compare with an uninfected unit of platelets. What media were the platelets cultured on? If it were Haemophilus (a short Gram negative rod), blood agar wouldn't be sufficient. There are other fastidious organisms as well, though the responding microbiologist is unaware of them being isolated from platelet concentrates. Some fastidious organisms (Strep. pneumo. being the classic example) can grow to high concentrations in liquid media and then autolyse and become nonviable. Autolysed pneumococci can appear gram negative. The observed symptoms would be a reaction to the bacterial by-products in the unit, not to viable bugs. It might be possible to work several of these options up by bacterial-universal-primer 16S PCR/sequencing if that's available, though it's hard to say how stable the DNA is in this sort of setting.

4. A colleague from Baltimore wrote that data she presented at the 2001 AABB meeting in abstract form (S Campbell-Lee, J Dick, T Kickler, PM Ness. Microbial Decontamination of Platelet Concentrates Using High Dose Gamma Irradiation ) might answer the question raised by the inquiring blood banker. According to the Baltimore blood banker "Our experiment was designed to examine whether gamma irradiation could be used to prevent bacterial contamination in stored platelet concentrates. After one milliliter of bacteria (S. epidermidis or Escherichia coli) suspended in saline was added to the platelets to achieve a final concentration of 10,000,1000, or 100 bacteria per milliliter,each platelet product was then immediately split into six bags and was given one of the following doses of irradiation using a blood bank irradiator: none, 100,200,300,400 or 500 Gy. A 0.5 ml sample was taken after irradiation and cultured. At 10,000 or 1000 bacteria per milliliter, there were >100 colonies after each dose of gamma irradiation. At 100 bacteria per milliliter, which represents a concentration of bacteria more likely seen in contamination cases, growth of either type of bacteria was not significantly decreased until a radiation dosage of 400 Gy was used. There were an average of 31 colonies of S. epi and 4 colonies of E. coli that were able to grow after this level of irradiation. Gram negative rods appear to be more sensitive to gamma irradiation, but at the dosage used for prevention of GVHD, it is unlikely that the growth of any contaminating bacteria was inhibited.

5. A blood banker from Pittsburg cautioned that bacterial contamination of the Gram stain reagents is not uncommon. He reports having seen this problem before. In addition, he points out that the dose of irradiation delivered to platelets prior to transfusion is not sufficient to prevent bacterial growth.

Editor's NOTE: This is an opportune time for members to review the CDC's BaCon study along with the report of results by Kuehnert MJ et al., Transfusion, Dec., 2001.

ADDENDA Mar. 18, 2002

6. Two blood bankers from Louisiana State University Health Sciences Center report that in 1996 they conducted an experiment to see if the currently required dose of gamma irradiation (i.e. 25 Gy) used in a blood bank would inhibit the growth of a bacterium, S. epidermidis in platelet units. The data showed that there was an average of 71% reduction of bacteria in the Irradiated Group over Non-Irradiated Group. From these data, they concluded that the conventional gamma irradiation currently recommended for blood components can NOT sterilize bacteria that may be in platelet units (Stradley KS & Kao, YS. The Louisiana blood bankers point out that surgical equipment, such as disposable scalpels, are sterilized with 25 kGy; which is 1,000 times of the dosage given to blood products.

7. The laboratory that performed the microbiology testing for the case report described by the inquiring blood banker has the following additional information in response to some of the comments that have been provided up to this time: "We certainly had many experienced microbiologists examine the Gram stain. The culture technique that we used should have grown Haemophilus and other Gram negative rods easily. Contamination of Gram stain reagents is always a possibility, although with our volume of work, we turn over the reagents very fast, making this less likely.

8. A blood banker from the East Coast said that she would investigate the possibility that the platelet pool showed Gram negative rods due to the contamination of the stain reagent. The responding blood banker admits that she had had patients with similar clinical circumstances to those described who actually had a septic transfusion reaction. In fact, she comments that just 4 days ago her lab had a pool of irradiated and leukocyte reduced platelets that grew out Gram positive organisms. The recipient of this pool of platelets was a child with aplastic anemia, who was on no antibiotics at the time, and whose blood cultures grew out the same organism. In the past when the reporting blood banker has had these kinds of cases and reported them to her blood supplier, the blood supplier quarantined all derived products and took cultures from the segments, so that they invariably got no growth, due to the minimal volumes cultured in the old-fashioned systems that they used. They do not use a Microtrak. They then tell the responding blood banker that they observe their nurses and there is no breach of practice.

Please submit comments to the e-Network Forum.

Ira A. Shulman, MD
CBBS e-Network Forum Editor & Moderator