The Effect of Donor Packed Cell Volume and Storage Characteristics on Transfusion Efficacy in Cats

Abstract

Purpose

To determine whether donor packed cell volume (PCV) and storage-related changes in feline packed red blood cell (pRBC) units influence transfusion efficacy in cats.

Patients and Methods

In this prospective, observational study, 35 client-owned, transfusion-naïve, anemic cats that received a single pRBC transfusion during hospitalization were enrolled. Donor unit characteristics including age, PCV, total solids, potassium, and ammonia were measured prior to transfusion. Transfusion response was assessed as the change in recipient PCV, normalized to transfusion dose (mL/kg), measured immediately post-transfusion and at 12 and 24 hours post-transfusion. Transfusion efficacy was assessed as the change in recipient PCV at each timepoint.

Results

Dose-normalized response was available for 34 cats immediately post-transfusion, 30 cats at 12 hours, and 16 cats at 24 hours. Median recipient pre-transfusion PCV (17%) increased significantly at all post-transfusion timepoints (T0: 21%, T12: 23%, T24: 24%; p≤,0.002). Increasing donor unit age was associated with higher ammonia concentrations and lower donor unit PCV but not with potassium concentration. No significant associations were identified between donor unit characteristics and transfusion efficacy.

Conclusion

Transfusion of pRBCs in cats resulted in rapid and sustained increase in recipient PCV. No association was detected between donor unit PCV or storage-related characteristics and the magnitude of this response. These findings suggest that transfusion response is multifactorial and may reflect a combination of donor-, product-, and recipient-related factors.

Introduction

Red blood cell (RBC) transfusion is a common therapeutic intervention in veterinary emergency and critical care. Packed red blood cell (pRBC) administration is indicated to improve tissue oxygen delivery in patients with anemia secondary to blood loss, hemolysis, and/or impaired RBC production. However, the clinical response to transfusion is variable and unpredictable and additional evidence-based dosing is needed. Several human and veterinary studies have assessed recipient-, donor-, and product-related factors that may contribute to variability in transfusion efficacy and patient outcomes. These include transfusion trigger selection, donor-recipient compatibility, and blood product storage duration.^1,2 Although the ideal transfusion trigger in people has not been determined, no consistent mortality benefit has been demonstrated between liberal versus more restrictive transfusion strategies (transfusing only when hemoglobin of <7-8g/dL).^3,4 Similarly, in cats, higher pretransfusion packed cell volumes (PCVs) (potentially reflecting too liberal of a transfusion trigger) have been associated with worse outcomes, but ideal transfusion thresholds in this species remain undefined.^1,5 Across species, individualized approaches to transfusion that incorporate underlying disease process, rate and rapidity of decline in hematocrit, and presence of clinical signs referable to anemia are generally recommended.^3,4

Optimization of donor-recipient compatibility through crossmatching has been proposed as a strategy to reduce transfusion-related complications and improve transfusion efficacy, however, results in cats are inconsistent.^6–8 Additionally, administration of older blood products is associated with worse outcomes in both people and small animals.^9–13 Storage-related biochemical and structural changes, collectively termed the RBC storage lesion, may impair RBC deformability, promote hemolysis, and reduce oxygen carrying capacity, limiting transfusion efficacy.^9,14 During storage, ammonia and potassium accumulate from cellular metabolism and efflux, respectively. Potassium accumulation may be less pronounced in stored feline pRBC units as feline erythrocytes have minimal Na⁺/K⁺-ATPase activity and consequently low intracellular potassium concentrations.^15,16 Donor-specific factors may also contribute to variability in transfusion response but have been minimally investigated in veterinary patients. The primary objective of the current study was to investigate the influence of donor unit PCV on transfusion efficacy in cats. Transfusion efficacy was defined as the dose-normalized change in recipient PCV following pRBC transfusion. An additional objective was to assess the effects of various RBC storage lesion parameters, including donor unit potassium and ammonia concentrations, on transfusion response. The study evaluated whether a higher donor unit PCV was associated with greater increases in recipient PCV and whether storage-related changes adversely impacted transfusion efficacy.

Given the potential functional limitations of stored blood products, optimizing transfusion efficacy is clinically relevant. Improved efficacy may reduce the need for repeated transfusions, thereby limiting cumulative exposure to the RBC storage lesion and decreasing the risk of transfusion-associated adverse events such as hypersensitivity reactions, transfusion-related lung injury, transfusion-associated circulatory overload, and infectious disease transmission. Reduced blood exposure to allogenic blood products may also translate to improved patient outcomes.

Materials and Methods

This prospective, observational study was conducted at the Schwarzman Animal Medical Center between December 2021 and April 2024. Thirty-five client-owned cats that received pRBC transfusions were enrolled as a single observational cohort. Cats were included if they were transfusion naïve, were administered a pRBC transfusion, and had PCV measurements obtained prior to transfusion and at standardized post-transfusion timepoints. Cats were excluded if the patient had received a prior transfusion, if transfusion volume could not be accurately determined, or if no post-transfusion PCV measurements were not collected. Given the small sample size, subjects missing data were included in analyses where able. Investigators and clinical personnel were not blinded to patient or blood product characteristics in this observational study. All investigations and treatments were performed with owner consent and were approved by the internal institutional animal care and use committee of the Caspary Research Institute of the Schwarzman Animal Medical Center. The authors confirm compliance with the journal’s ethical guidelines and adherence to best practices of veterinary care.

For each enrolled patient, collected data included signalment, body weight, blood type determined via blood typing card (DMS Laboratories, Inc., Temple, AZ) and crossmatch data. When performed, major crossmatching was conducted by trained personnel using the laboratory’s standardized procedure in which donor RBCs were combined with recipient serum and evaluated for evidence of incompatibility. Crossmatches were assessed for both macroscopic and microscopic agglutination. The presence of agglutination or hemolysis were considered indicative of incompatibility. The cause or suspected cause of anemia was also recorded. Transfusion-specific data collected included total transfusion volume (mL) and calculated transfusion dose (mL/kg), duration of transfusion, and number of transfusions administered during hospitalization.

Donor unit characteristics measured prior to transfusion included unit age (defined as time from collection to transfusion), PCV, total solids (TS), ammonia and potassium concentrations (Catalyst One, IDEXX Laboratories, Inc., Westbrook, ME). Packed cell volume was determined by microhematocrit centrifugation, and total solids was measured from plasma using refractometry. Recipient PCV was measured immediately prior to transfusion (baseline), immediately following transfusion (T0), and then at 12 hours (T12) and 24 hours (T24) following transfusion. Fluid therapy was administered at the discretion of the attending clinician based on patient needs and was not standardized as part of the study protocol. Concurrent fluid administration was not incorporated into assessments of PCV.

Data was assessed for normality using the Shapiro–Wilk test. Parametric data are reported as mean ± standard deviation, and non-parametric data are reported as median and range. To assess transfusion efficacy, PCV measurements were normalized to transfusion dose (mL/kg). To determine the increase in PCV per unit of transfused blood, the ΔPCV was calculated as the difference between the post-transfusion and pre-transfusion PCV divided by the transfusion dose (mL/kg) as follows:

The new, calculated parameters are referred to ΔPCV0, ΔPCV12, and ΔPCV24 to represent the three different time points assessed. Comparisons between continuous parameters over time were performed by Wilcoxon signed rank test. Forward, stepwise linear regression analyses were applied to assess for associations between measured variables. The model and goodness of fit were evaluated by likelihood ratio and comparison to chi-square assessment and all values evaluated as statistically significant demonstrated appropriate models. Bonferroni correction was utilized to control for multiple comparisons. Statistical significance was set at a P value of < 0.05. All statistical analyses were performed by commercial statistical software (StataSE 15.1, StataCorp LLC, College Station, TX).

Results

Thirty-five cats met the inclusion criteria and were enrolled in the study. The study population consisted of one intact male, 21 castrated males, 13 spayed female, and no intact female cats. The most highly represented breed was domestic shorthair (24/35), and the remaining breeds included a single participant each. The median age was nine years (range, 0.75–20 years) and mean body weight was 4.2 ± 1.3 kg. Most cats were blood type A (34/35) with one cat classified as blood type AB. All cats received type-specific transfusions. Major crossmatching was performed in 13/35 (37.1%) cats prior to transfusion, of which 11/13 (84.6%) were reported as compatible and 2/13 (15.2%) as incompatible. The specific cause of crossmatch incompatibility in these two cats was not determined. The remaining cats did not undergo pretransfusion crossmatching. The most common indication for transfusion was decreased erythropoiesis (13/35), followed by mixed etiology (combination of blood loss and decreased production [7/35]), blood loss (6/35), unknown cause (5/35), and hemolysis (4/35).

The median transfused volume was 29.7 mL (range, 19.6–41.0) and the median dose of transfusion was 7.0 mL/kg (range, 2.7–15). Twenty-nine cats were administered a single transfusion over the course of hospitalization, and six received a second transfusion at some point during hospitalization, which was not included in analysis.

Donor unit characteristics were available for most transfusions. Table 1 summarizes the characteristics of the donor population. Linear regression identified significant associations between donor unit age and both unit PCV and ammonia concentration. The mean unit age was 12.6 ± 6.6 days and increasing unit age was significantly associated with higher ammonia concentration (279.5 ± 109.4 µmol/L, 95% CI: 0.02–0.06) and with lower donor unit PCV (48.8 ± 5.7%, 95% CI: −0.8-⁻0.1, p=0.0002, R²=0.44; n=34). No significant associations were found between unit age and potassium concentration (3.5 ± 0.8 mmol/L, 95% CI: −2.9–3.7) or with total solids (3.6 ± 0.6 g/dL, 95% CI: −3.6–4.1).

Table 1.

Donor PRBC Unit Summary Characteristics Sample Size, Mean and Standard Deviation for Unit Age, Packed Cell Volume, Total Solids, Ammonia and Potassium for Stored PRBC Units

	Sample Size	Mean	Standard Deviation
Unit age (days)	34	12.6	6.6
Donor PCV (%)	34	48.8	5.7
Donor TS (g/dL)	34	3.6	0.6
Donor Ammonia (µmol/L)	33	279.5	109.4
Donor Potassium (mmol/L)	33	3.5	0.8

Abbreviations: RBC, red blood cell; pRBC pack red blood cells; PCV, packed cell volume.

The median pre-transfusion PCV in cats enrolled in the study was 17% (range, 8–25%). Post-transfusion PCV measurements were available for all 34 cats immediately following transfusion. By 12 hours, three cats had been discharged prior to PCV measurement and one cat did not have a PCV rechecked, resulting in 12-hour PCV measurements for 20 cats. By 24 hours, PCV measurements were available for 16 cats as 11 additional cats were discharged, one died, one was euthanized, and one did not have a 24-hour PCV rechecked. Recipient PCV increased significantly following transfusion at all timepoints (T0: 21%, range 15–37, p <0.001; T12: 23%, range 15–37, p<0.001; T24: 24%, range 14–32, p=0.002). No differences were observed between any post-transfusion PCV measures in the study. Mean dose-normalized changes were ΔPCV0 (0.86 ± 0.7%/mL/kg, n=34), ΔPCV12 (1.1 ± 1.2%/mL/kg, n=30), and ΔPCV24 (0.95 ±0.6%/mL/kg, n=16). No significant associations were identified between ΔPCV at any time point and donor unit PCV, ammonia concentration, potassium concentration, or unit age.

Discussion

This study evaluated the impact of feline donor unit PCV and storage-related variables, including unit age, potassium, and ammonia concentrations, on transfusion efficacy. Optimizing pRBC transfusion efficacy has the potential to reduce patient exposure to blood products and improve patient outcomes. Although transfusion resulted in a significant increase in recipient PCV, after normalization for transfusion dose, no association was detected between the donor unit PCV or the evaluated storage-related variables and the magnitude of this increase. These findings suggest that, within the range of donor unit PCVs evaluated in this study, transfusion response may reflect the combined influence of recipient-, product-, and donor-related factors.

In human medicine, donor-related characteristics have been proposed to influence transfusion outcomes, although evidence remains limited and inconsistent.¹⁷ It has been reported that pre-donation hemoglobin levels in human blood donors are associated with reduced pRBC unit quality and decreased unit hemoglobin content.¹⁸ However, studies have failed to demonstrate a consistent relationship between donor characteristics and post-transfusion changes in recipient hematocrit or PCV.¹⁷ Human blood banking practices aim to standardize unit PCV at 55–65%, theoretically resulting in predictable increases in PCV (~1-3%) following transfusion.¹⁹ These estimations are based on oversimplified calculations that do not account for variability in donor unit PCV, transfusion volume per body weight dose, or recipient physiology.²⁰

Similarly, in veterinary medicine, commonly cited transfusion formulas aim to standardize dosing and predict post-transfusion PCV, with the general expectation that administration of 1 mL/kg of pRBCs will increase PCV by approximately 1%.¹⁹ Two commonly applied formulas in small animals include:

• (a)Volume of pRBCs (mL) = 1.5 x desired increase in PCV (%) x body weight (kg)
• (b)Volume of pRBCs (mL) = ([desired post-transfusion PCV – patient PCV] / donor unit PCV) x blood volume (mL/kg) x body weight (kg)^20,21

Although formula (b) incorporates donor unit PCV, neither formula performs well in predicting post-transfusion response.^8,21 Prior veterinary studies report lower than expected increases in PCV following transfusion in cats, with similar responses observed in both crossmatched (0.75%/mL/kg) and uncrossmatched (0.62%/mL/kg) patients, suggesting that incompatibility alone does not explain this variability.⁸ Additionally, a large retrospective study in cats found no difference in PCV increase following administration of whole blood versus pRBCs, despite differences in red cell concentration.¹⁹ These findings suggest that transfusion response in cats may not be reliably predicted by donor unit PCV alone.

Additional intrinsic donor variables, including sex, differences in red cell enzyme activity, susceptibility to oxidative stress, and membrane stability, have been proposed to affect transfusion efficacy in human studies.²¹ While analogous donor-specific variables have not been clearly defined in small animal medicine, it is reasonable to consider that feline erythrocytes may have greater susceptibility to oxidative stress due to higher numbers of sulfhydryl groups, Heinz bodies, and a greater abundance of oxidizable thiol residues on hemoglobin compared to other species, which may contribute to variability in transfusion response in this species.^5,22

Consistent with prior studies, storage-related changes were evident in donor units in the present study. Increasing unit age was associated with higher ammonia concentrations and lower donor unit PCV, likely reflecting ongoing metabolic degradation and RBC structural changes.^16,23 Ammonia accumulation in stored RBC units results from the breakdown of adenine nucleotides and intracellular proteins and has been proposed as a contributor to adverse transfusion events, rather than as a determinant of transfusion efficacy.^23,24 The observed decrease in donor unit PCV with age is likely due to increased time for RBC metabolism and hemolysis.

Despite measurable changes in storage-related variables, no association was identified between these factors and transfusion efficacy in the present study. Similar findings are reported in human medicine, with older RBC units demonstrating increased hemolysis and reduced hematocrit without consistent adverse effects on transfusion outcomes.¹⁸ These data support the continued use of available feline pRBC units within accepted storage guidelines without expectation of reduced efficacy based solely on unit age. However, these findings should not be interpreted as evidence that donor age or storage-related characteristics are clinically unimportant under all circumstances. Rather, donor- and product-related characteristics should be considered in conjunction with recipient-specific clinical factors when selecting blood products.

In contrast to what is observed in stored human blood, potassium did not increase with storage in feline pRBC units. In people, potassium accumulation occurs as declining ATP levels impair the function of ATP-dependent RBC Na⁺/K⁺-ATPase.^25,26 In cats and dogs, however, erythrocytes have low intracellular potassium concentration and lack Na⁺/K⁺-ATPase activity, which likely explains the absence of this storage-related change.²⁷

Independent of donor unit characteristics, recipient-related variables likely also play a substantial role in determining transfusion response. Continued blood loss or erythrocyte destruction may attenuate the expected increase in PCV following transfusion, and concurrent fluid administration may decrease measured PCV through hemodilution. Inflammation may further contribute to changes in PCV through decreased RBC survival, iron sequestration, and impaired erythropoiesis, as demonstrated in cats with naturally occurring inflammatory disease.²⁸ These factors are particularly relevant in the heterogenous population evaluated in the present study, in which these clinical variables could not be controlled for and may have contributed to variability in post-transfusion PCV. This may have obscured associations between donor unit characteristics and transfusion efficacy.

In addition, the role of pre-transfusion crossmatching in cats remains complex and incompletely defined.^6–8 Crossmatch testing may improve donor-recipient compatibility by identifying antibodies to non-AB blood group antigens, including Mik, and may reduce the risk of transfusion reactions and potentially improve efficacy; however, its impact on post-transfusion PCV is inconsistent.^6–8 Some studies report greater increases in post-transfusion PCV with crossmatch-compatible transfusions, while others do not, underscoring the lack of clear, evidence-based recommendations for routine crossmatching in transfusion naïve cats.^1,6–8

All cats in this study received type-specific transfusions, and a subset underwent pre-transfusion crossmatching. Among those tested, the majority were crossmatch compatible. A small number of cats received crossmatch-incompatible transfusions. Given the limited number of crossmatched cases, formal comparative analyses between groups was not performed. Crossmatching is generally not uniformly performed prior to initial transfusion in clinical practice, and its impact on transfusion outcomes remains uncertain.

In the present study, transfusion efficacy was defined as the magnitude of increase in post-transfusion PCV, a commonly used surrogate for improved oxygen-carrying capacity through increased circulating hemoglobin. However, PCV alone does not reflect the physiologic impact of transfusion, including effects on oxygen delivery, microcirculatory function, and red blood cell survival. Direct evaluation of RBC survival and function following transfusion would likely provide a more accurate measure of transfusion benefit, but such tests are difficult to perform clinically and require specialized equipment and advanced methodologies. Techniques such as near-infrared spectroscopy and specialized magnetic resonance imaging can assess tissue and cerebral oxygen extraction, while biotin labeling of RBCs allows for evaluation of post-transfusion RBC recovery and survival kinetics.²⁹ These approaches remain largely experimental and require further investigation to determine their utility in veterinary medicine.

There are several limitations to this study. An a priori sample size calculation was not performed and the relatively small sample size, significant attrition (mostly due to patient discharge), and narrow range of donor unit PCVs may have limited the statistical power to detect subtle associations. This may have resulted in false-negative findings and attrition bias. Additionally, transfusion efficacy was assessed solely based on PCV change, and other clinically relevant outcomes, including transfusion reactions, inflammatory response, and patient survival, were not evaluated. Potential confounding factors such as crossmatch compatibility, disease severity and inflammation, continued blood loss and concurrent fluid therapy, were also not controlled for.

Despite these limitations, pRBC transfusion resulted in a significant and sustained increase in PCV at all assessed post-transfusion timepoints, with stabilization occurring within 24 hours. These findings support the clinical utility of pRBC transfusion in anemic cats, however, donor- or storage-related variables predictive of transfusion response were not identified.

Conclusion

In this study, no association was detected between donor unit PCV or the storage-related variables evaluated and transfusion efficacy as measured by increase in recipient PCV. Although storage duration was associated with higher ammonia concentrations and lower donor unit PCV, these changes did not translate into clinically relevant differences in post-transfusion PCV at any timepoint. Although no association was detected between unit age and transfusion efficacy, progressive ammonia accumulation during storage must be considered when selecting feline pRBCs, particularly for patients in which an ammonia load may be clinically relevant, such as those with hepatic encephalopathy. Transfusion efficacy is likely multifactorial and influenced by a combination of donor-, product-, and recipient-related factors. Future studies incorporating larger, multicenter populations and functional measures of transfusion efficacy may help further define the relative contributions of various factors to transfusion response.

Disclosure

Dr Alicia Mastrocco is a handling editor for Journal of Veterinary Emergency and Critical Care, outside the submitted work. The author(s) report no conflicts of interest in this work.