Fatal Sepsis from Bacterial Contamination of Transfusion Platelets

On June 29, 2018, the CDC published a report that described two separate clusters of platelet transfusion–associated bacterial sepsis in Utah and California during August 2017. Four of five transfused patients died, though the cause of one of those deaths was not definitively the platelet transfusion–associated bacterial sepsis. Investigations of the blood suppliers and the hospitals revealed no procedural deviations. Findings in this report highlighted that even when following current procedures, the risk for transfusion-related infection and fatality persists, making additional interventions necessary.


In Utah, two patients died after platelet transfusions from the same donation. Clostridium perfringens isolates from one patient’s blood, the other patient’s platelet bag, and donor skin swabs were highly related by whole genome sequencing (WGS).


Two apheresis platelet units and one unit of plasma were manufactured from an apheresis blood donation in Utah. Both platelet units were distributed to a hospital.


A male (patient A) with acute myeloid leukemia and neutropenia received one of the platelet units. Thirty minutes after transfusion, patient A developed rigors, a sudden feeling of cold with shivering accompanied by a rise in temperature, often with copious sweating, especially at the onset or height of a fever. Transfusion-transmitted bacterial infection was not considered then because of the patient’s complex medical history. The patient died four days later. Anaerobic blood cultures, obtained shortly after transfusion, grew C. perfringens five days after collection.


Fourteen hours after patient A’s transfusion, a female (patient B) with acute myeloid leukemia received the other platelet unit while on broad-spectrum antibiotics for neutropenia at the same hospital. No immediate symptoms of sepsis followed transfusion. Later that day, routine laboratory testing revealed new intravascular hemolysis, the abnormal breakdown of red blood cells in the blood vessels. Transfusion-transmitted bacterial infection was suspected. Gram stain of platelet bag residuals was performed, revealing gram-positive bacilli. The platelet supplier was immediately notified. Patient B died 11 hours after transfusion. C. perfringens was isolated from an anaerobic culture of the residual platelets. Post-transfusion blood cultures from patient B were negative.


Platelet units transfused to patients A and B had been collected four days before transfusion. Routine inoculation for aerobic culture, performed 24 hours after donation, was negative for bacterial growth through five days.


The donor had previously donated platelets and whole blood with no recipient adverse reactions reported. The health department interviewed the donor, who reported no relevant infectious exposures or symptoms. The donor consented to skin swabs. Consent for environmental sampling was not provided by the donor.


As part of the investigation, multiple samples from the donor, recipients, and platelet bags were cultured for C. perfringens under anaerobic conditions. DNA was isolated from cultures that had growth (donor swabs, patient A’s blood, two isolates of patient B’s platelet bag residual, and one control [an unrelated C. perfringens isolate]). WGS indicated all six epidemiologically linked isolates were highly related.


An investigation of the blood supplier and the hospital revealed no procedural deviations. The non-transfused plasma unit from the donor was quarantined. The donor was permanently deferred.


In California, one patient died after a platelet transfusion. Klebsiella pneumoniae isolates from the patient’s blood and platelet bag residuals and a non-transfused platelet unit were matched using WGS. An investigation of the blood supplier and the hospital revealed no procedural deviations.


Three apheresis platelet units and one unit of plasma were manufactured from an apheresis blood donation in California. One platelet unit was distributed to one hospital, where it was divided into two portions, known as aliquots. Two platelet units were distributed to a second hospital.


At the first hospital, one platelet aliquot was transfused to a male (patient C) with myelodysplastic syndrome, fever, and neutropenia, who was on multiple broad-spectrum antibiotics. Approximately nine hours after transfusion, the patient developed septic shock but recovered. Multiple post-transfusion blood cultures were negative, presumably a result of the antibiotic regimen.


Five hours after patient C’s transfusion, the second aliquot was transfused to a male (patient D) who had received an autologous stem cell transplant. Patient D developed vomiting, tachycardia, and hypotension approximately 15 minutes after transfusion initiation. Despite discontinuing transfusion, patient D died within five hours. Multiple post-transfusion blood cultures drawn after the transfusion reaction grew K. pneumoniae. Transfusion-transmitted bacterial infection was suspected. Gram stain of platelet bag residuals was performed, revealing gram-negative rods. The blood supplier was immediately notified. K. pneumoniae was isolated from the platelet bag residuals.


At the second hospital, one platelet unit was transfused to a female (patient E) with disseminated intravascular coagulation and septic shock, for which patient E was receiving broad-spectrum antibiotics. Patient E died the following day. Blood cultures obtained at the onset of sepsis (pre-transfusion) and eight hours after transfusion both grew multidrug-resistant K. pneumoniae.


One day after patient E’s transfusion, the blood supplier notified the second hospital that gram-negative rods had been identified in the residual aliquot transfused into patient D. The hospital returned the non-transfused platelet unit from which K. pneumoniae was later isolated.


The routine donor’s platelet bacterial screening collection, inoculated 24 hours after donation, was negative for growth through five days. The frozen plasma unit was not cultured and was discarded. K. pneumoniae isolates from three patient D blood cultures, patient D’s residual platelet product, and the second hospital’s non-transfused platelets had similar antibiograms and were highly related by WGS. However, pre-transfusion and post-transfusion K. pneumoniae isolates from patient E demonstrated multidrug resistance and were unrelated from the other isolates using WGS. Patient E’s possible source of sepsis was a pre-transfusion urine infection with multidrug-resistant K. pneumoniae.


Investigation of the blood supplier and hospitals indicated no procedural deviations. The donor met eligibility criteria and frequently donated platelets, but had been deferred multiple times because of low hemoglobin. A platelet donation nine months earlier was positive for Enterobacter cloacae. After the report of the K. pneumoniae cluster, medical history assessments did not identify donor bacterial infection risks. Non-transfused blood products from the implicated donation were quarantined, and the donor was permanently deferred.


Platelet-transmitted bacterial infections persist as a cause of transfusion-associated morbidity and mortality. Contamination of blood products most commonly occurs when skin microbiota are introduced during needle insertion, but can also occur from asymptomatic donor bacteremia. Because the majority of platelets are stored at room temperature, bacteria can proliferate to clinically important levels by the time the unit is transfused. Approximately one in 5,000 platelet collections are contaminated with bacteria, and one in 100,000 platelet transfusions results in bacterial sepsis. Transfusion-transmitted bacterial infections are likely underdiagnosed because recipients are often given broad spectrum antibiotics, like patient C; have underlying medical conditions that increase sepsis risk, like patient A; or the septic reaction might not be attributed to the transfusion.


Current practices to mitigate the risk for bacterial contamination of platelets include donor health screening, skin examination and disinfection, diversion of up to the first 40 mL of blood into a separate non-transfusable pouch to reduce the introduction of skin flora, visual inspection of platelet bags before transfusion, and aerobic bacterial culture screening (e.g., monitoring an aliquot for bacterial growth) at least 24 hours after platelet collection. Investigations confirmed that the Utah and California collection facilities followed current practices. This report highlights that, even when following current practices, the risk for fatalities persists, making additional, important interventions necessary.


The FDA has several recommendations related to platelet contamination and donation. The FDA recommends that blood suppliers control the risk for bacterial contamination either by using a pathogen reduction device or performing bacterial detection at least once. Additional requirements when a pathogen is identified include product quarantine, organism identification, determination whether the pathogen is endogenous to the donor blood stream, and, if so, donor deferral.


Additional evidence-based risk mitigation strategies, including pathogen inactivation, rapid detection at point-of-use, and modification of screening bacterial culture protocols, can reduce the risk for platelet-transmitted bacterial sepsis. Implementation of these modified and alternative strategies in the United States has been supported by advice from the FDA’s Blood Products Advisory Committee, but are not currently required. Pathogen inactivation technology was adopted in France, Belgium, and Switzerland, and although no confirmed septic transfusion reactions were reported from 2.3 million pathogen inactivation–treated platelet units, two possible cases have been reported after transfusion of pathogen inactivation–treated platelets. This same pathogen inactivation technology is approved by the FDA for use with apheresis platelets and plasma in the United States.


Rapid bacterial detection devices, optimally used 72 hours after collection, can detect bacteria using less than one milliliter of platelet volume, but only have detection limits of 103–106 organisms/mL. The FDA has cleared one rapid device for extending platelet shelf life from five days to seven days.


Additional risk mitigation strategies modify existing bacterial culture screening protocols. Current methods differ by blood supplier, with most inoculating 8 mL into an aerobic blood culture microbial detection system sampled 24 or more hours after collection to allow for sufficient bacterial growth. If cultures are negative after 12 to 24 hours, platelet units are released and have a shelf life of up to five days, which can be extended up to seven days with secondary testing. However, 8 mL of platelets sampled 24 hours after donation might not have sufficient bacterial loads to detect bacterial growth in the screening culture. Rather than using a fixed volume, one proposed strategy involves using a minimal proportional sample volume of 3.8 percent of the platelet total collection. For example, in the United Kingdom, culture volumes of 16 mL are divided equally between aerobic and anaerobic culture bottles 36 to 48 hours after donation and have resulted in no recognized fatalities after approximately 1.8 million platelet units were transfused with shelf life extended to seven days. However, on several reported occasions, platelet bags were suspected of contamination after visual inspection, and subsequent cultures confirmed contamination. In Ireland, repeat aerobic and anaerobic bacterial cultures are performed four days after collection to extend platelet shelf life to seven days. No septic transfusion reactions have been reported after more than 100,000 apheresis collections. Although reporting by blood systems that have adopted modified culture screening methods is promising, demonstrating important clinical benefit is difficult because transfusion-associated bacterial sepsis is rare. However, when compared with current detection practices in the United States, methods based on larger volume culture (as in the UK), delayed sampling of platelets (as in Ireland), and performing aerobic and anaerobic cultures after collection (as in the UK and Ireland) are likely to result in fewer cases of platelet-transmitted bacterial infections.


C. perfringens, a sporogenic gram-positive bacterium, has been rarely reported as the source of transfusion-associated sepsis. Disinfectants used for skin antisepsis during blood collection are not sporicidal and might be ineffective in removing C. perfringens from skin. K. pneumoniae, a gram-negative bacterium, is a common pathogen among transfusion-related fatalities. Both pathogens might not be inactivated by pathogen inactivation, but might have been detected with the modified culture strategies described above, which are not routinely practiced in the United States.


Blood collection services should consider implementing enhanced safety interventions to reduce further the risk for bacterial contamination of platelets. Clinicians should consider bacterial contamination when patients develop sepsis during or after a platelet transfusion and rapidly investigate these transfusion reactions.


See the CDC Report


See also Medical Risk Law Report: Blood Draws, Testing, Transfusions: Venipuncture Injury, Inaccurate Results, Tainted Blood - The Liability Risks


See also Medical Risk Law Report: How Risky Is Going to the Hospital? The Dangers and Liabilities of Healthcare-Associated Infections


See also Medical Risk Law Report: Hospital-Acquired Infections: Who Is Liable and Why?