Differential effects of speed and volume on transfusion‐associated circulatory overload: A randomized study in rats

Abstract Background and Objectives Transfusion‐associated circulatory overload (TACO) is the primary cause of transfusion‐related mortality. Speed and volume of transfusion are major risk factors. The aim of this study was to investigate the interaction of red blood cell (RBC) transfusion speed and volume on the development of TACO. Materials and Methods A validated model for TACO in anaemic Lewis rats with an acute myocardial infarction was used. The effect on pulmonary hydrostatic pressure of one, two or four units of packed RBCs transfused in either 30 or 60 min was evaluated (3.3–26.6 ml·kg−1·hr−1). Pulmonary capillary pressure was measured as left ventricular end‐diastolic pressure (LVEDP). Cardiac stress biomarkers atrial natriuretic‐peptide (ANP) and N‐terminal pro‐brain natriuretic peptide (NT‐proBNP) were measured 1‐h post‐transfusion. Results Thirty animals were included (n = 5 per group). Transfusion of RBCs increased LVEDP in a volume‐dependent manner (ΔLVEDP [mmHg]: −0.95, +0.50, +6.26, p < 0.001). Fast transfusion increased overall ΔLVEDP by +3.5 mmHg and up to +11.8 mmHg in the four units' group (p = 0.016). Doubling transfusion speed increased ΔLVEDP more than doubling volume in the larger volume groups. No difference in ANP or NT‐proBNP were seen in high transfusion volume or groups. Conclusion Transfusion volume dose‐dependently increased LVEDP, with speed of transfusion rapidly elevating LVEDP at higher transfusion volumes. ANP and NT‐proBNP were not impacted by transfusion volume or speed in this model. TACO is seen as purely volume overload, however, this study emphasizes that limiting transfusion speed, as a modifiable risk factor, might aid in preventing TACO.


INTRODUCTION
Transfusion-associated circulatory overload (TACO) is the leading cause of transfusion-related respiratory distress, ICU admission and death [1][2][3]. The incidence of TACO in those transfused is 1% of in-hospital patients [4,5], and up to 6% in the ICU [6,7]. Both the speed and volume of transfusion are associated with TACO [5,6,8].
Even though millions of transfusions are carried out each year, surprisingly few studies address transfusion speed or volume.
TACO is the result of hydrostatic pulmonary oedema, where transudate is driven into alveoli restricting gas exchange [9,10]. Both speed of transfusion and large volumes acutely increase circulating volume which can back-up the pulmonary circulation and increase pulmonary capillary pressure (P cap ) [11]. Our previously published animal model has shown that after an equivalent volume, blood transfusion increases P cap more than crystalloids [12]. Similarly, in ICU patients transfusion increases P cap and is associated with increased mortality [13]. Only two studies in humans have investigated P cap related to transfusion speed, which increases in a speed-dependent fashion [14,15], transfusion volume however has not previously been investigated.
Investigating modifiable risk factors including speed and volume effects of transfusion individually as well as their interaction is necessary to understand TACO. Measuring P cap is difficult and can differ throughout the pulmonary capillary bed, therefore, left-atrial pressure (LAP -the pressure downstream) is used in clinical research. In humans, this requires a Swan-Ganz catheter to estimate LAP which is invasive and not without risk [16]. In animals direct intra-cardiac measurement of left ventricular end-diastolic pressure (LVEDP), which is equal to the maximum LAP , can be investigated through [12].
The study aimed to investigate the relative contribution of transfusion speed and volume on the development of TACO.
Atrial natriuretic peptide (ANP) and N-terminal pro-brain natriuretic peptide (NT-proBNP) will be evaluated as volume overload biomarkers. We hypothesize that both a rapid speed of infusion, as well as larger transfused volumes will increase LVEDP and cardiac stress biomarkers. Using a previously validated rat model of TACO, animals will be randomized and the effect of one, two or four units of packed red blood cells (RBC) transfused in either 30 or 60 min on LVEDP will be measured.  [17]. Animals were housed in a specialized animal care facility, exposed to standard light-dark cycle and fed standard rat chow with ad-libitum access to water.

MATERIALS AND METHODS
A two-hit model for TACO was used as previously published [12], which is a clinically relevant anaemia-transfusion model in a mechanically ventilated, cardiac comprised setting ( Figure 1). Two hits are required for TACO to develop as healthy anaemic animals were able to accommodate large transfusion volumes. Anaemia was induced by replacing blood with an equivalent volume of colloids, maintaining circulating blood volume (CBV). Isovolemia is important as intravascular hypovolemic patients will be able to accommodate volume, evidenced by the highest incidence of TACO in normovolemic patients with chronic anaemia [18,19] and INR correction through plasma transfusion in non-bleeding patients [20]. The first hit is a patient risk factor, F I G U R E 1 Experiment design comparing speed versus volume of transfusion. A two-hit model for TACO in anaemic rats was employed. The first hit, an MI results in volume incompliance. Animals were randomized to speed of transfusion (30 or 60 min) and further randomized to receive one, two or four unitsa total of six groups. Art.line, Arterial cannula; CVC, Central venous cannula; LAD, Left anterior descending coronary artery; PV-Catheter, Pressure-volume (catheter) resulting in volume-incompliance, in this case myocardial dysfunction [21]. This lowers the threshold to develop TACO following a second hita blood transfusion.

Animal procedures
General anaesthesia was induced in healthy animals through brief use of isoflurane (5%) followed by an intra-peritoneal injection of racemic ketamine (9 mgÁ100 g À1 ), dexmedetomidine (12.5 μgÁ100 g À1 ) and atropine (5 μgÁ100 g À1 ). Ketamine was continuously infused through a tail-vein cannula to maintain anaesthesia (5 mgÁ100 g À1 Áh À1 ), chosen for its limited cardiovascular depressive profile. Animals were ventilated via a tracheostomy secured with a suture around the trachea.

Cardiac catheterization and line placement
An ultraminiature rat pressure-volume catheter (SPR-838, Millar, USA) was passed through the right carotid artery into the left-ventricle.
Calibrations of the PV-catheter include: a two-point pressure calibration, blood conductivity calibration using the cuvette method and correction for parallel conductance using hypertonic saline boluses (NaCl 30%, 10 μl bolus) [22]. Conductivity calibrations were performed at fixed timepoints following changes in blood electrolyte, that is: (1) baseline, (2)  Isovolemic anaemia and first hit -Myocardial ischemia Approximately 20% of CBV was replaced over a period of 15 min with an equivalent volume of colloid solution (Tetraspan 6%, B. Braun, Germany) until a haematocrit of 30% was achieved. CBV (ml) was calculated as 6.5% of body weight (g) [23]. Afterwards animals were allowed to stabilize for 15 min.
A myocardial infarction (MI) was induced as first hit. A continuous norepinephrine infusion was started and titrated to a MAP of 65 mmHg. Through a left-anterior thoracotomy (AE2 cm incision through the fourth intercostal rib) the left-anterior descending artery (LAD) was permanently ligated using a 5-0 Prolene suture [12,24]. An intrapleural chest drain (20G) was placed one rib below the incision and the thorax and skin were closed in two layers using a 3-0 Vicryl suture. After closing the chest, the drain was removed under negative pressure while simultaneously a recruitment manoeuvre was performed to minimize residual air. Animals were allowed to stabilize for 30 min.

Randomization and second hit -Transfusion
A sealed envelope, block randomization method was used to stratify animals to either one, two, or four units to be transfused using a volumetric pump, performed in equal groups per week (a 1:1:1 ratio).
Animals were then again randomized by sealed envelope to rapid or slow transfusion (over 30 or 60 min). From randomization onward, no further alterations were allowed to ventilator settings, fluid, norepinephrine or anaesthetic infusion rates. Volume of RBC units was predetermined and fixed (1.0 ml per unit) calculated as 5% of CBVequivalent to 80 kg male humans. Hemodynamic parameters were recorded continuously until the experiment ended at 1-h posttransfusion and the rats were terminated. The RBC units were buffy coat reduced packed red blood cells products, manufactured following Dutch blood banking procedures, made from the whole blood

Sample size
Sample size was determined using previously published data [12], and further extrapolated assuming a linear correlation between ΔLVEDP T A B L E 1 Characteristics pre-transfusion

RESULTS
A total of 30 animals were included, randomized to one, two or four units transfused (n = 10 per group) and further randomized to either slow or rapid infusion (n = 5 per group). Prior to randomization seven animals died during cannulation due to cardiac tamponade (n = 2), during thoracotomy (n = 1) or due to arrhythmia's following myocardial infarction (n = 4). Animals that died prior to randomization were replaced; no animals died after randomization.
The effects of the first hit were seen as (1)   significantly associated with an increase in ΔLVEDP only in the group transfused with four units (Figure 2).

Changes in LVEDP between volume transfused and speed of infusion
A secondary analysis was performed to determine the contribution of speed compared to transfused volume on LVEDP. Both doubling of speed and volume of transfusion increased ΔLVEDP (Table 2).
At low volumes the changes were not-clinically relevant, however at larger volumes, a doubling of transfusion speed resulted in an exponential increase ΔLVEDP (four-times higher).

Changes in haemodynamics following transfusion
With increasing transfused volume and increased preload, seen as increased LVEDP, no increase in overall CO was detected (Table S1,

Pulmonary outcomes
Overall, no differences were found in pulmonary outcomes (Table S2).
P/F-ratio at termination nor wet-dry ratio or histopathological examination of the lungs (where pulmonary oedema was ranked on a 0-3-point scale by an experienced pathologist) differed between the number of units transfused, nor the speed of transfusion.

Biomarkers
Overall, ANP and NT-proBNP were elevated (Figure 3), however they were not significantly difference in between the number of units An increased LAP is the mechanism behind developing TACO.
A landmark study in dogs showed that hydrostatic pulmonary oedema started developing at pressures above 25 mmHg [26]. There is limited evidence of how speed and volume individually contribute to the development of TACO. Mechanistic transfusion studies have been performed in both healthy dogs [27] and swine [28]. Both show an increase in pulmonary pressures following transfusion, similar to our results. Two studies in humans [15,18] showed a transfusion speeddependent increase, also similar to our model. Our model is the first to employ packed red blood cells contrary to use of whole blood.
Rheological properties of units with 60% haematocrit are likely to alter hemodynamic outcomes unpredictably. Second, this is the first study to investigate both transfused volume and speed. There is a clear stepwise increase in LVEDP as result of increasing transfusion volume. Transfusion speed seems particularly important to higher transfusion volumes doubling of speed increased LVEDP exponentially (Figure 1 and Table 2).
Our data suggesting transfusion speed is a critical factor is an important signal to clinicians that could prevent TACO.
Successful interventions to reduce over-transfusion include patient blood management programs, transfusion triggers and a single-unit transfusion in non-emergency situations [29]. Splitting an RBC unit can also reduce the volume infused over time, allowing transfusion over days without exposing a patient to multiple donors. Finally, volume-reduced RBC units, hyperconcentrated platelets and partially reconstituted lyophilized F I G U R E 3 Volume overload biomarkers. Data presented in a Tukey boxplot plasma have been suggested to reduce the volume load of transfusion [9,30]. Reducing transfusion speed in non-emergency settings is an easy intervention, even so, there is likely room for improvement as retrospectively in 50% of TACO cases infusion speed was not specified [31]. There is limited guidance on speed of transfusion, aside from a 4-h maximum to prevent bacterial growth. Noted is that transfusion speed may be higher if tolerated by the patients circulatory system [32]. Specifically, in at-risk patients guidance regarding the speed of transfusion is warranted.

Cardiac overload biomarkers
While markers are elevated, ANP and NT-proBNP levels did not contribute in discerning different volumes or speeds of transfusion ( Figure 3).
ANP is used experimentally and is secreted from preformed vesicles and the atria, ready for rapid release [33]. NT-proBNP is widely used in clini- increased LAP as well as increased mortality [13].
We utilized a specialized rat model involving an acute MI and transfusion. This model provides a mechanistic approach to circulatory overload and simplifies how TACO develops in humans, who are often older and have different predisposing conditions. We have previously shown that transfusion increases LVEDP not only after an MI but also with underlying acute kidney injury (AKI) [14]. Therefore transfusion of RBC's increases P cap in volume incompliant rats independent of cardiac function, which broadens the generalizability of our results using an MI model.
Interestingly a recent RCT found very low incidences of acute heart failure following liberal transfusion (3.7%) versus restrictive transfusion (3.2%, n.s.) [34]. Unfortunately, the study did not specifically score TACO, nor was the speed of transfusion documented. The low incidence of TACO may additionally be explained by other interventions such as routine echocardiograms and two-thirds of patients were either already using or started on diuretics within 24-h of admission, all of which being interesting targets for TACO research.
Translating our findings will inevitably require humans studies to guide transfusion speed. While slower is likely better, not all patients will require this, and an individual risk-assessment should be performed. Future studies using this model should also focus on interventions to prevent and treat TACO, for example through loop-diuretics or volume-reduced blood products. The upcoming challenge will be the development of a TACO phenotype model with evident pulmonary oedema, in which interventions can start targeting pulmonary oedema. Future studies in humans should make sure transfusion speed is equal between groups. This study paves the way to a study in high-risk patients (i.e., non-bleeding patients >60-years-of-age with either pre-existing renal or cardiac dysfunction) to be randomized to receive either a split or full RBC unit over 1 or 2 h.
In conclusion, this study demonstrates a stepwise transfusion volume-dependent increase of LVEDP which is key in the development of TACO. Transfusion speed appears critical in developing circulatory overload, specifically when larger volumes were transfused. Cardiac biomarkers ANP and NT-proBNP were not correlated with the volume or transfusion speed 1-h after transfusion. This study underlines a policy of slow transfusion in non-urgent cases.