Novel approaches to measure transfusion effectiveness

Abstract

Purpose of review:

This review encompasses different considerations of transfusion effectiveness based upon clinical scenario and transfusion indication. Tissue oxygenation, cerebral metabolic oxygen use, and red blood cell (RBC) survival are important elements of transfusion effectiveness in individuals with acute and chronic transfusion requirements.

Recent findings:

Non-invasive measures of tissue and cerebral oxygen extraction include near infrared spectroscopy (NIRS) and specialized magnetic resonance imaging (MRI) sequences. RBC survival timepoints including 24-hour post-transfusion recovery, 50% recovery timepoint, and mean potential lifespan may be accurately measured with biotin-labeling of RBC prior to transfusion. Labeling at different cell surface densities allows survival of multiple RBC populations to be determined.

Summary:

While past trials of optimal transfusion thresholds have focused on Hb as a singular marker for transfusion needs, measures of oxygenation (via NIRS or specialized MRI) and RBC survival (via biotin labeling) provide the opportunity to personalize transfusion decisions to individual patient’s acute health needs or chronic transfusion goals.

INTRODUCTION

The goals of transfusion are fundamentally defined by the clinical scenario in which transfusion is indicated. In situations such as acute hemorrhage or critical illness, red blood cell (RBC) transfusion is indicated to immediately improve oxygen delivery to tissues and brain and improve survival.¹ Chronic transfusion typically is employed for patients with long-term transfusion dependence due to congenital anemias such as sickle cell disease (SCD) or transfusion-dependent thalassemia (TDT), acquired myelodysplastic syndromes, or malignancies. With chronic anemias, RBC transfusion goals typically focus on maintaining a target hemoglobin (Hb) goal, persistence of adequate Hb levels post-transfusion, and suppression of endogenous erythropoiesis. These goals of chronic transfusions are balanced against aims of limiting cumulative transfusion donor exposures and iron overload. Unique to SCD, the hematologic goal of transfusions is the dilution or reduction of circulating hemoglobin S (HbS)-containing erythrocytes, to reduce acute sickle-related complications and/or progression of vasculopathy.^2–4 Consequently, post-transfusion donor RBC survival important to achieving effective transfusion therapy in the chronic setting.

Measures of RBC transfusion effectiveness must reflect the specific clinical and hematologic goals for the clinical scenario. In settings of acute illness, surgery, or hemorrhage, immediate Hb increment, 24-hour post-transfusion recovery (PTR-24) survival measurements, and brain and tissue oxygenation are important measures of transfusion effectiveness. In chronic transfusion therapy, Hb increment and rate of clearance between transfusions directly impact transfusion intervals and number of exposures over time. The kinetics of in vivo RBC survival impact therefore is an important determinant of transfusion effectiveness.⁵ Direct measurement of RBC survival may be achieved by taking advantage of genetic differences between the donor and recipient RBC (e.g. donor/recipient minor RBC antigen differences or HbA and HbS quantification in recipients with SCD),⁶ or through labeling techniques such as chromium or biotin. Effective transfusion in all settings depends on donor-recipient compatibility and a lack of hemolytic transfusion reactions. While transfusion compatibility is predicted in vitro by routine crossmatch or occasionally the monocyte monolayer assay (in complex alloimmunization scenarios), direct measurements of in vivo donor RBC survival provide post-transfusion confirmation of the rate of clearance and true compatibility of the transfusion.

Tissue and cerebral oxygenation measures

In acute transfusion, rise in Hb is the simplest means of judging transfusion outcome, yet controversy exists surrounding what level of anemia warrants transfusion in different clinical scenarios and whether liberal or restrictive transfusion thresholds are superior. Physiologic measures of anemia tolerance and oxygenation may be more appropriate to guide transfusion management. In critically ill patients with increased tissue metabolism, tissue oxygen consumption (VO₂) and delivery (DO₂) may be calculated from cardiac output and arterial and venous oxygen concentrations, allowing for determination of tissue oxygen extraction ratio before and after transfusion.⁷

Near infrared spectroscopy (NIRS) or Diffuse Optical Spectroscopy (DOS) is a non-invasive technique that uses the absorption and reflection of near-infrared light wavelengths to measure regional oxygen saturation (difference between oxyhemoglobin and deoxyhemoglobin) in specific tissues.^8–10 NIRS is commonly used in neonates to monitor cerebral and gut oxygenation and may be used to study the impact of RBC transfusion on neonatal outcomes such as necrotizing enterocolitis and to identify individualized transfusion triggers.¹¹ NIRS/DOS allows the measurement of cerebral oxygenation by light emission and absorption at the tissue surface, while diffuse correlation spectroscopy (DCS) allows for measurement of cerebral blood flow (CBF) by measuring the dynamic scattering of infrared light by circulating RBCs.¹² Both techniques are restrained by depth sensitivity and signal limits, particularly at the skull where extracranial blood flow measurements may interfere; however they provide non-invasive means to assess hemo-metabolic stress and response to transfusion. These techniques have been used in acute settings to monitor regional cerebral oxygen saturation (crO₂) and assess the impact of transfusion in in critically ill neonates, congenital heart disease, and children receiving extra-corporal membrane oxygenation (ECMO)^13–15 and have significant clinical research applications in the chronic transfusion setting, particularly for stroke prevention in SCD.

In SCD, a hematologic goal of chronic transfusion typically is maintaining HbS <30%. However, progressive cerebral infarcts still occur at unacceptably high frequency even among patients who achieve this target HbS levels.¹⁶ SCD is associated with decreased arterial oxygen content, which leads to compensatory abnormally high cerebral oxygen extraction fraction (OEF) and CBF to meet cerebral metabolic rate of oxygen utilization (CMRO₂). Increased CBF and OEF have been shown to predispose individuals with SCD to stroke and are reduced following RBC transfusion, thereby reducing cerebral metabolic demands that predispose to ischemic injury.¹⁷ Frequency-domain NIRS and DCS have been used in a study of 35 chronic transfusion events in children with SCD to show significant decreases in CBF index, cerebral blood volume, and cerebral OEF immediately after a single transfusion, without change in CMRO₂.¹⁸ Novel magnetic resonance imaging (MRI) sequences, specifically arterial spin labeling (ASL) for CBF and asymmetric spin echo for OEF quantification may also be used to assess the cerebroprotective goals of chronic transfusions in SCD. Utilizing ASL-MRI before and after RBC exchange, Guilliams et al. demonstrated significant decreases in CBF and OEF while maintaining CMRO₂ in 21 children with SCD receiving chronic transfusions for stroke prevention. Further, regions of peak OEF in the deep white matter (a common area of SCD-related silent infarcts) were identified and noted to also show improvement with transfusion.¹⁹ These non-invasive imaging techniques may assist in identifying those patients with SCD undergoing chronic transfusion who remain at higher stroke risk based on CBF and OEF, and thus allow clinicians to modify transfusion to individuals’ specific cerebroprotective needs.

HbA increment for measuring transfusion survival in sickle cell disease

Transfusion in SCD presents the unique opportunity to distinguish donor RBCs from autologous RBCs through Hb separation (via electrophoresis) and quantification of the percentage of circulating donor RBCs (containing HbA). In chronically transfused patients with SCD, the change in HbA concentration (total Hb (g/dL) x HbA fraction (%)) between transfusion events can be calculated as the difference in HbA immediately post-transfusion from HbA prior to the next transfusion.²⁰ In this instance, change in HbA concentration (ΔHbA) serves as a surrogate for the rate of clearance of donor RBCs, and has been used to study donor and recipient factors influencing the rate of transfused RBC clearance. In a cohort of 63 children with SCD on chronic transfusion therapy, ΔHbA was calculated for each transfusion event over 12 months to assess for differences in HbA clearance. A high rate of variability in ΔHbA was seen among different patients as well as between transfusion events within the same patient. Increased ΔHbA was also associated with lower total Hb, higher HbS, and higher reticulocyte count at the next transfusion, demonstrating that ΔHbA predicts effectiveness of erythropoiesis suppression. ΔHbA was greater in patients with past RBC alloimmunization, lower in patients with splenectomy, and greater following transfusion of units with severe G6PD deficiency.²¹

Using this surrogate marker of donor RBC clearance, a significant association of G6PD deficiency with Duffy-null phenotype in donor units suggests that Duffy-null antigen selection of RBC may also select for donors with RBC traits common to African and Middle Eastern ancestries which may impact long term (> 24 hours) clearance of donor RBCs.²²

In acute settings, ΔHbA has great utility in assessing for accelerated clearance of donor RBCs post-transfusion, aiding in the diagnosis of delayed hemolytic transfusion reactions (DHTRs) in SCD. DHTRs remain an under-recognized transfusion complication in individuals with SCD due to challenging diagnostic criteria, and their tendency to mimic vaso-occlusive crises. While drop in total Hb and evidence of increased hemolysis may be related to the underlying pathophysiology of SCD, the rate of HbA decline provides more specific assessment of donor RBC clearance. Mekontso-Dessap et al. developed a diagnostic nomogram comparing HbA within 1 week of transfusion to a later point post-transfusion to assess the likelihood of DHTR based on ΔHbA and the number of days post-transfusion. Through use of routine Hb electrophoresis, this nomogram allows an initial predictive assessment to prompt further investigation for and timely treatment of DHTRs in individuals with SCD.^23,24

RBC Labeling for Direct Post-Transfusion Survival measurements

While HbA is a useful marker of transfusion, its application is largely limited to transfusions in individuals with SCD. It also is imprecise, as it may reflect multiple different RBC units transfused at multiple occasions within the past 120 days or more. To understand the survival of one unique recent transfusion episode or multiple different RBC units, precise RBC labeling is required. The gold standard for evaluating RBC recovery has been chromium-51 (⁵¹Cr) labeling. PTR-24 determination by ⁵¹Cr-RBC has been a regulatory criteria of the Food and Drug Administration (FDA) for licensing of new RBC storage systems or preparation processes for more than four decades.⁶ However, ⁵¹Cr labeling carries many disadvantages. Use of radioactive ⁵¹Cr isotope requires laboratory oversight and raises ethical concerns for ⁵¹Cr exposure of vulnerable patients. Additionally, ⁵¹Cr prohibits tagging multiple RBC populations simultaneously, and is unable to reliably measure RBC survival past approximately 35 days. Chromium tags RBCs though noncovalent binding to Hb, and eluates at a rate of about 1% per day, although individual rate has been shown to be more variable in some hematologic disorders, limiting the ability to obtain accurate survival measurements over days to weeks.²⁵ Beyond its use for regulatory requirements, chromium-labeling is impractical for more routine use in the clinical or clinical research setting.

Biotin labeling of RBCs has no toxicity or radioactivity, provides sensitive in vivo tracking for the full duration of RBC survival due to stable covalent bonding to cell surface proteins, and offers the ability to label multiple RBC populations simultaneously with discrete biotin densities (table 1). The biotinylation agent N-hydroxysulfosuccinimide-biotin (sulfo-NHS-biotin) covalently bonds biotin to primary amine groups found on lysine side-chains, allowing for labeling of RBC surface proteins (figure 1a). RBCs collected from an autologous or allogeneic donor are washed and labeled with biotin at different densities, allowing them to be distinguished separately by flow cytometry with streptavidin. RBCs from different sources (allogeneic and/or autologous) may be biotinylated at different surface densities, and the post-transfusion survival of different RBC populations followed simultaneously within the recipient (figure 1b) down to levels of 0.06% of total circulating RBC.²⁶ With stable binding of the biotin label, time points beyond PTR-24 can be reliably assessed, including the RBC population’s half-life (T₅₀), percentage survival at various time points of interest, and estimation of mean potential lifespan (MPL).

Table 1.

Comparison of methods of RBC tagging to measure post-transfusion survival kinetics

	Chromium	Biotin
Pros	Standardized, historically accepted gold standard method PTR-24 method required by FDA for licensing of new RBC systems	Non-toxic, No radioactivity Stable covalent bond with amino groups on RBC surface proteins, does not elute off RBC during lifespan Accurate and sensitive detection of very dilute populations of RBC (down to 0.06%) to measure the full RBC lifespan Ability to label RBC with different densities of surface biotin, clearly distinguishable by flow cytometry.
Cons	Radioactivity exposure to patient Laboratory regulation and oversight required for radioactive handling Non-covalent bond, decays at unpredictable rate from Hb Short term survival studies only Inability to distinctly label 2 or more different RBC populations	Potential immunization to biotinylated RBC, limiting future studies Time sensitive process, requiring sterile handing of an open RBC product Research use: Investigational new drug approval (or local regulatory oversight) required

PTR-24: 24-hour post-transfusion recovery

Figure 1. RBC Biotinylation and Survival Measurements.

Figure 1a. A biotinylating agent such as N-Hydroxysulfosuccinimide biotin (sulfo-NHS-biotin) reacts with primary amino (-NH₂) groups on cell surface proteins of washed RBC, forming stable covalent amide bonds.

Figure 1b. Representative example of a patient with SCD who received a 3 unit biotin-labeled allogeneic RBC transfusion. RBC are selectively labeled with different surface densities of biotin, making them easily distinguishable by streptavidin flow cytometry. At serial time points post-transfusion, the percentage of biotin-RBC is multiplied by total RBC count and compared to immediate post-transfusion measurements. Subsequent monthly transfusions will dilute the percent of remaining biotin-RBC.

Antibodies to Biotin-labeled Red Blood Cells

In autologous biotin-RBC survival studies of health volunteers, approximately 15% of participants demonstrated antibodies to biotin-RBC at ≥12 weeks following initial exposure, with no evidence of accelerated clearance or hemolysis following the initial immunizing event.²⁷ In subsequent re-exposures to autologous biotin-RBCs in subjects with biotin-RBC immunization, an anamnestic biotin-RBC antibody response (IgG₁ subclass) occurred within 5–7 days. While no clinical adverse events were noted, accelerated clearance of biotin-RBCs was seen with more rapid clearance of biotin-RBCs with higher biotin density and with repeated subsequent exposures.²⁸ Biotinylation for RBC survival studies therefore requires pre-exposure screening and close monitoring post-transfusion for biotin-RBC antibodies. Biotin-RBC antibodies were only noted in individuals who received biotin densities >18 µg/mL, with accelerated clearance in immunized individuals correlating with higher biotin density. Therefore, it is recommended to select labeling densities and volumes to limit overall density x volume dose to ≤180 µg per exposure.²⁸

Biotin-RBC Uses

RBC biotinylation has been used in a variety of clinical scenarios to assess both autologous and allogeneic donor RBC survival. Autologous studies in healthy adult volunteers have helped define normal RBC survival kinetics, measure red cell volume, and study the impact of biotin-RBC antibodies and clinically significant antibodies (e.g. anti-D) on RBC survival.^26,29,30 Initial autologous studies used up to 4 simultaneous labeling densities (6–162 µg/mL), finding similar RBC survival curves for densities ≤54 µg/mL, while RBC labeled at higher densities (>54 µg/mL) showed more rapid clearance, possibly due RBC membrane damage.^27,28 Given this and biotin-RBC antibody considerations, contemporary biotin studies should use the lowest achievable biotin densities that can still be distinguished from unlabeled and other labeled RBC populations by flow.

Autologous RBC biotinylation may also be used to study RBC survival in chronic hematologic diseases such as SCD. The differential survival of RBCs based on high or low levels of fetal Hb, and the understanding of different sickle RBC populations derives from autologous collection, biotinylation, and re-infusion of RBCs for full lifespan studies.^31–33

Both autologous and allogeneic RBC biotinylation has been used in studies of anemia, blood volume, and RBC lifespan in preterm and very low birth weight infants. Total red cell and blood volumes, which determine oxygenation and transfusion needs, have been studied through biotinylation and re-infusion of very small (≤0.5 mL) autologous blood volumes.^34,35 Volume measurements with RBC biotinylation have also been used to study the impact of delayed cord clamping in preterm infants.³⁶ Multiple density biotin studies of autologous and allogeneic RBCs in infants have shown shortened T₅₀, and MPL of neonatal RBCs as compared to donor RBCs, but also shorter MPL of allogeneic donor RBCs within the neonate.^37–39

Allogeneic RBC survival studies using biotinylation in chronically transfusion in SCD have begun to assess donor and recipient characteristics influencing RBC survival. Evaluating survival kinetics of 2 RBC units in 9 patients with SCD and 5 with thalassemia, from the time of transfusion until their next transfusion, Gerritsma et al. demonstrated inter-patient variability in survival to be more pronounced than intra-patient variability.⁴⁰

The impact of RBC donor variability and storage conditions may also be assessed with biotinylation studies, with greater accuracy compared to chromium labeling. Storage age has been assessed by collecting and biotinylating autologous RBCs at 2 distinct densities, reinfusing one labeled group after 5–7 days storage and the second group at 35–42 days of storage, and comparing the clearance kinetics of the 2 RBC populations. While showing no significant differences in T₅₀ nor MPL, this study provides a mathematical model for calculating full RBC lifespan and highlights the advantage of biotin to compare different RBC unit characteristics simultaneously.⁴¹

With allogeneic transfusion, multiple donor factors may be assessed simultaneously with biotin-labeling of 2 or more transfused RBC populations. Clinical applications of biotin in allogeneic transfusion also include assessing RBC clearance rates in patients with RBC autoantibodies or other antibodies of uncertain clinical significance, providing in vivo assessment of transfusion compatibility and effectiveness. Research measurements of immunologic response or markers of RBC senescence such as phosphatidlylserine externalization could be correlated with exact survival of transfused biotin-labeled RBCs.⁴²

CONCLUSION

Transfusion decision making based upon Hb transfusion triggers have historically resulted in conflicting outcomes and controversy over transfusion thresholds and choice of outcome measurements. With individual assessment of transfusion goals such as tissue oxygenation status, metabolic needs, or chronic transfusion requirements, decisions can be based upon novel measures other than Hb. Non-invasive measures such as NIRS and MRI offer new ways to determine RBC transfusion needs and metabolic response to transfusion. Biotin labeling allows for improved evaluations of transfusion efficacy and mechanisms of different hemolytic states. Coupled together, these two technologies may offer novel insights into transfusion efficacy in myriad of patient populations and clinical scenarios.

KEY POINTS.

1. Transfusion effectiveness is defined by short and long-term goals of transfusion in settings of acute anemia, critical illness, or chronic hematologic disease.

2. Tissue and cerebral oxygen consumption and extraction may be measured by non-invasive measures such as NIRS and MRI to determine RBC transfusion needs and metabolic response to transfusion.

3. RBC survival has an important role in the effectiveness of chronic transfusion therapy and may be safely and directly measured with biotin labeling of RBC.