Deadly Bacteria in Donated Platelets: Two 2017 Clusters in Utah and California

Bacteria in platelet transfusions still sicken and kill patients. In August 2017, two unrelated clusters of sepsis after platelet transfusions were reported, one in Utah and one in California. Three patients died. Investigators found no departures from procedure at the blood suppliers or the hospitals, and the report concludes that, since the risk persists even then, more safeguards are needed.

Utah: Clostridium perfringens

One apheresis donation in Utah yielded two platelet units and one unit of plasma. Both platelet units went to hospital X, where they were transfused four days after collection. The routine aerobic culture, started 24 hours after the donation, had shown no growth for 5 days.

• Patient A, a man with acute myeloid leukemia and neutropenia, developed rigors 30 minutes after his transfusion. Because his medical history was complicated, an infection from the transfusion was not suspected then. He died 4 days later. Blood drawn soon after the transfusion grew C. perfringens in anaerobic culture 5 days after it was collected.
• Patient B, a woman with acute myeloid leukemia who was on broad-spectrum antibiotics for neutropenia, received the second unit 14 hours later. She had no immediate signs of sepsis, but routine tests that day showed new intravascular hemolysis. Staff suspected contaminated platelets, stained the leftover product, saw gram-positive bacilli and alerted the supplier at once. She died 11 hours after the transfusion. C. perfringens grew from the leftover platelets, though her own blood cultures after the transfusion were negative.

The donor had given platelets and whole blood before without any reported reactions, and told the health department of no relevant infections or exposures. Skin swabs from the donor grew C. perfringens. Whole genome sequencing showed that the six linked isolates — from the donor's skin, patient A's blood and patient B's platelet bag — were almost identical to each other and clearly different from an unrelated control. The donor's plasma unit was quarantined, and the donor was permanently deferred.

Timeline of the Utah and California clusters. CDC, MMWR.

California: Klebsiella pneumoniae

One apheresis donation in California produced three platelet units and a plasma unit. Hospital Y received one unit and split it in two; hospital Z received the other two. The routine culture, again taken 24 hours after donation, showed no growth for 5 days.

• Patient C, a man who had had an autologous stem cell transplant, developed vomiting, tachycardia and hypotension about 15 minutes into his transfusion. The transfusion was stopped, but he died within 5 hours. His blood cultures and the leftover platelets grew K. pneumoniae.
• Patient D, a man with myelodysplastic syndrome, fever and neutropenia who was on several broad-spectrum antibiotics, had received the other half of that unit 5 hours earlier. About 9 hours later he went into septic shock, but he recovered; his blood cultures were negative, probably because of the antibiotics.
• Hospital Z returned its untransfused unit, which grew K. pneumoniae. Its other unit had already gone to patient E, a woman with septic shock and disseminated intravascular coagulation, who died the next day. But her infection turned out to be a different, multidrug-resistant strain, present before the transfusion and probably from a urinary infection.

Sequencing showed that the K. pneumoniae from patient C and from the two platelet products were highly related, differing by only two single-nucleotide changes. The donor met all eligibility criteria and gave platelets often, though they had been turned away several times for low hemoglobin, and a platelet donation 9 months earlier had tested positive for Enterobacter cloacae. No infection risk was found in the donor's history. The remaining products were quarantined and the donor permanently deferred.

Why the safeguards missed it

Blood products are most often contaminated by skin bacteria carried in when the needle goes in, but a donor can also carry bacteria in the blood without symptoms. Because most platelets are stored at room temperature, bacteria can multiply to dangerous levels before the unit is used. About one in 5,000 platelet collections is contaminated, and about one in 100,000 platelet transfusions causes sepsis. The true number is probably higher, because recipients are often on antibiotics or already at risk of sepsis, and a reaction may never be blamed on the transfusion.

Current safeguards include screening donors' health, examining and disinfecting their skin, diverting up to the first 40 mL of blood into a separate pouch, inspecting bags before transfusion, and an aerobic culture at least 24 hours after collection. Both collection centers had followed all of them. The two organisms also posed particular problems: skin disinfectants do not kill the spores of C. perfringens, which has rarely been reported in transfusion sepsis, while K. pneumoniae is a common culprit in transfusion deaths. Both might not be inactivated by pathogen inactivation, but both might have been caught by the modified culture methods used elsewhere.

Stronger options

The Food and Drug Administration requires blood suppliers to control bacterial risk either with a pathogen reduction device or by testing for bacteria at least once, and sets out what must happen when a pathogen is found — quarantine, identification, and deferral if it came from the donor's bloodstream. The FDA's Blood Products Advisory Committee has supported further measures, which are not required:

• Pathogen inactivation. Used in France, Belgium and Switzerland; among 2.3 million treated platelet units there were no confirmed septic reactions, though two possible cases have been reported. The same technology is approved in the United States for apheresis platelets and plasma.
• Rapid tests at the point of use. Best done 72 hours after collection, they need less than 1 mL of platelets but detect bacteria only at 10³–10⁶ organisms per mL. The FDA has cleared one device for extending platelet shelf life from 5 to 7 days.
• Better culture screening. Most U.S. suppliers culture 8 mL, sampled at least 24 hours after collection, which may not hold enough bacteria to detect. Alternatives include sampling a set proportion (3.8%) of the collection; the United Kingdom's method of 16 mL split between aerobic and anaerobic bottles 36–48 hours after donation, with no recognized deaths among about 1.8 million units given with a 7-day shelf life; and Ireland's repeat aerobic and anaerobic cultures on day 4, with no septic reactions reported after more than 100,000 apheresis collections.

Because transfusion sepsis is rare, proving the benefit of these methods is hard, but larger, later and anaerobic as well as aerobic cultures are likely to prevent more infections than current U.S. practice. Clinicians should suspect contamination when a patient develops sepsis during or after a platelet transfusion, investigate quickly, and report reactions to blood suppliers and hemovigilance systems.

Sources

• Horth RZ, Jones JM, Kim JJ, et al. "Fatal Sepsis Associated with Bacterial Contamination of Platelets — Utah and California, August 2017." MMWR 67(25). CDC.