Red cell physiologic stress results in lower quality transfusions: a randomized trial in adults with sickle cell disease

Abstract

People with sickle cell disease (SCD) may be transfused with red cell units that are near the end of their storage life, exposing them to components of the red cell storage lesion. This study evaluated the clinical impact of storage age and red cell distress markers on chronically transfused adults with SCD. This randomized prospective clinical trial recruited 26 chronically transfused adult patients (aged >16 years) with SCD; 13 participants were randomized to each study arm, that is, targeted to receive only ≥30-day or ≤10-day stored red cell units for 3 consecutive outpatient transfusion events. The red cell units were evaluated via quantification of surface exposure of phosphatidylserine (PS) and phosphatidylethanolamine (PE). Differences in key clinical variables were also evaluated. We show that patients receiving units with higher surface-exposed PS and PE, regardless of storage age, had a reduced hemoglobin (Hb) increment at 2 weeks (PS-PE high, 0.59 g/dL; PS-PE low, 1.04 g/dL; P = .04), increased pain crises (incidence rate ratio, 7.44; 95% confidence interval [CI], 1.53-36.25), and higher odds of reported illness (odds ratio, 1.94; 95% CI, 1.06-3.54). Furthermore, posttransfusion serum iron predicted subsequent subjective symptoms of illness (receiver operating characteristic-area under the curve, 0.76) and correlated with smaller increases in HbA percent after transfusion (r =-0.36; P = .04). These data suggest that the physiologic state of transfused red cells may deleteriously affect patient outcomes. Future studies focusing on identifying and then avoiding lower-quality red cell units in high-risk patients with SCD could enhance patient safety. This trial was registered at www.clinicaltrials.gov as #NCT03704922.

Figure 1.. Histogram of storage age of the units transfused by study arm (red, ≤10-day arm; blue, ≥30-day arm).

There were 8 transfused units (12.3%) stored for >10 days in the fresh unit arm, and 11 transfused units (16.4%) stored for <30 days in the old unit arm (P = .6).

Figure 2.. Distribution of surface-exposed PS (A) and PE (B) on stored RBCs.

Aliquots were evaluated by flow cytometry. There was no significant difference between the 2 arms regarding PS surface positivity, but units transfused in the ≥30-day arm had a significantly higher percent of PE-positive RBCs than those in the ≤10-day arm (P = .0004).

Figure 3.. Adjusted change in Hb (with 95% CI) and HbA percent per day from the pretransfusion value across all 3 transfusion events, stratified by phlebotomy time point and adjusted for the number of units transfused.

(A-B) Hb increment 2 hours after transfusion (34 pre-to-post deltas based on 14 unique participants), 24 hours after transfusion (29 pre-to-post deltas based on 14 unique participants), and 2 weeks after transfusion (55 pre-to-post deltas based on 24 unique participants) are shown by study arm (A) and by PS-PE surface content (B). No significant differences were noted by study arm at 2 hours or 2 weeks after transfusion but was significant at 24 hours after transfusion (2.22 vs 3.23 g/dL; P = .007). Transfusion episodes classified as PS-PE high had significantly smaller Hb increments at both 24 hours and 2 weeks after transfusion, compared to PS-PE mixed and low transfusion events (24 hours, 2.38 vs 2.82 vs 2.89 g/dL [P = .02]; 2 weeks, 0.59 vs 0.89. vs 1.04 g/dL [P = .04]). (C-D) The adjusted mean decrement in HbA percent per day (with 95% CI) across all 3 transfusion events during the first 2 weeks after transfusion, the second 2 weeks after transfusion, and overall (change between exchanges) are shown by study arm (C) and during the first 2 weeks and second 2 weeks post-transfusion by PS-PE surface content (D). Transfusion-specific decrement per day from 2 hours to 2 weeks after transfusion is based on 28 transfusions involving 13 unique patients. Transfusion-specific decrement per day from 2 weeks to before the next transfusion included 19 total assessments from 11 unique patients. The overall change in HbA percent per day by study arm was –0.24% for those receiving ≤10-day units and –0.14% for those receiving ≥30-day units (P = .3). There were no significant differences in the change in HbA percent per day by study arm or PS-PE, but PS-PE high and ≥30-day units had greater decrements in the first 2 weeks.

Figure 4.. Scatterplot demonstrating the negative correlation (best fit regression lines are shown) between the peak serum iron at 2 hours after transfusion and the decrease in Hb (A) and HbA percent decrement after transfusion (B).

Although the negative correlation was seen at all time points observed, the negative correlation was significant for Hb at 2 weeks after transfusion (n = 32; r = –0.42; P = .01) and HbA percent increment at 2 hours (n = 32; r = –0.4; P = .03) and 24 hours (n = 28; r = –0.5; P = .01) after transfusion.

Figure 5.. Patient-reported symptoms of illness during the study interval.

Seven of the 26 patients reported at least 1 subjective symptom of possible illness during the study interval, accounting for 45 of the 1434 diary days (3.1%) recorded (A). Forty of those 45 days (88%) were after transfusion with PS-PE high units (4.0% of diary days; 103 involved transfused units; P = .004), and no symptoms were noted in those who received PS-PE–low units (0% of diary days and 27 involved transfused units). The odds of having subjective symptoms reported during the physical examination after transfusion with PS-PE–high units was 1.94 (95% CI, 1.06–3.54). No statistically significant differences were observed when comparing study arms (P =.1; odds ratio, 1.41; 95% CI, 0.25–7.87). (B) The most common symptoms noted were upper respiratory in nature (ie, rhinitis, cough, and fatigue).