Is Intracerebral Hemorrhage Risk Shared via Blood Transfusion?

Recipients of red blood cell transfusions from donors who later had multiple intracerebral hemorrhages (ICHs) had a significantly increased risk of spontaneous ICH themselves, an exploratory retrospective cohort study showed.

Among over 1 million patients in two cohorts from Sweden and Denmark, those who underwent red blood cell transfusion from donors who developed multiple spontaneous ICHs had a significantly higher risk of a single spontaneous ICH compared with patients receiving transfusions from donors who did not develop spontaneous ICH:

• Swedish cohort: unadjusted incidence rate 3.16 vs 1.12 per 1,000 person-years (adjusted HR 2.73, 95% CI 1.72-4.35, P<0.001)
• Danish cohort: unadjusted incidence rate 2.82 vs 1.09 per 1,000 person-years (aHR 2.32, 95% CI 1.04-5.19, P=0.04)

No significant difference was observed for recipients of transfusions from donors who developed a single spontaneous ICH in the Swedish cohort (aHR 1.06, 95% CI 0.84-1.36, P=0.62) nor the Danish cohort (aHR 1.06, 95% CI 0.70-1.60, P=0.73), reported Jingcheng Zhao, MD, PhD, of Karolinska University Hospital Solna in Stockholm, and colleagues in JAMA.

"The observed increased risk of spontaneous ICH associated with receiving a red blood cell transfusion from a donor who later developed multiple spontaneous ICHs, corresponding to a 30-year cumulative incidence difference of 2.3%, is a novel finding," the group noted.

"This may suggest a transfusion-transmissible agent associated with some types of spontaneous ICH, although the findings may be susceptible to selection bias and residual confounding, and further research is needed to investigate if transfusion transmission of CAA [cerebral amyloid angiopathy] might explain this association," they concluded.

There were also no differences seen for ischemic stroke, which the researchers explained was used as a negative control outcome because it shares risk factors with spontaneous ICH, but is not associated with CAA.

Indeed, sensitivity analyses in the Swedish cohort showed no significant associations for a first ischemic stroke used as a negative control outcome (HR 1.12, 95% CI 0.79-1.58) or when using ischemic stroke both in blood donors as the exposure and in transfusion recipients as the outcome, among recipients of transfusions from donors who had a single stroke (HR 0.97, 95% CI 0.92-1.03), or among recipients of transfusions from donors who had multiple strokes (HR 0.88, 95% CI 0.76-1.02).

Although the study does not directly assess CAA, Zhao and team aimed to evaluate possible transfusion-related transmission of agents leading to CAA, identified as a top priority by a recent international consortium.

"We expect CAA to be more prevalent among donors who develop multiple spontaneous ICHs, as CAA-related ICH has been reported to have a 7-fold increase for recurrent ICHs compared with non-CAA-related ICH," they wrote.

In an accompanying editorial, Steven M. Greenberg, MD, PhD, of Harvard Medical School in Boston, noted that "there are good reasons to treat the possibility of CAA transmission via blood transfusion seriously -- and good reasons to remain skeptical at least for the present."

Greenberg added that he remains unconvinced due to "the weakness of evidence for a plausible biological mechanism for blood from a donor with future CAA to rapidly transmit CAA-related hemorrhage. The short-time course is quite challenging to explain: Nearly half of the ICHs among blood recipients occurred within 5 years of transfusion, dramatically faster than the 30- to 40-year interval reported between neurosurgical exposure to cadaveric tissue and first ICH."

According to a statement from the study authors, "this study does not demonstrate causality. The most important implication of the study is ... that it adds further support to the hypothesis that CAA can be transmitted between individuals, which, if true, can have consequences in several fields."

Although dementia was not assessed as an outcome, sensitivity analyses showed that recipients of transfusions from donors who later developed both a single spontaneous ICH and dementia were significantly more likely to develop a single spontaneous ICH (HR 2.44, 95% CI 1.52-3.94), which Zhao and co-authors noted could be interpreted in the context of CAA frequently co-occurring with Alzheimer's dementia.

For this study, the researchers used nationwide blood bank and health register data from Sweden (main cohort) and Denmark (validation cohort), including all 1,089,370 patients ages 5 to 80 years recorded to have received a red blood cell transfusion from January 1970 (Sweden) or January 1980 (Denmark) through December 2017.

Among the 759,858 patients in the Swedish cohort, median age was 65, and 59% were women, while among the 329,512 patients in the Danish cohort, median age was 64, and 58% were women. Median follow-up for the two cohorts was 5.8 years and 6.1 years, respectively.

Exposures of the two cohorts were equal, with 1.1% of each group receiving a blood unit from a donor who later developed a single spontaneous ICH and 0.1% of each group having a donor who later developed multiple spontaneous ICHs.

Comment