Plasma transfusions in horses with typhlocolitis/colitis

Abstract

The outcome of treatment of horses with plasma for typhlocolitis/colitis at the Ontario Veterinary College-Health Sciences Centre was evaluated. Horses with typhlocolitis/colitis that received a plasma transfusion had higher odds of dying than did non-transfused horses. The clinical usefulness of transfusing plasma to hospitalized hypoproteinemic horses is questioned.

Plasma transfusions are frequently used in equine medicine for various medical conditions (1). Patients with chronic or acute hypoproteinemia as a result of protein-losing enteropathies, strangulating lesions of the intestine, and enterocolitis, for example, are common candidates to receive plasma (1). Plasma transfusions have been advocated for one or more of the following: to restore or improve colloid osmotic pressure (COP) with the aim of preventing edema; to provide a source of protein for nutritional support; to counteract endotoxemia; and to supplement immunoregulatory factors such as opsonins, complement system, and immunoglobulins (1).

In equine medicine, clinical studies measuring the benefit or outcomes of such intervention in hypoproteinemic horses are lacking. Whilst guidelines for use of plasma transfusion are clearly defined in many texts (total protein < 40 g/L, albumin < 20 g/L, and COP < 12 mmHg), and transfusing plasma to hypoproteinemic horses appears intuitive, the rationale for its use is not evidence-based (1–3). Some prospective studies investigating the use of plasma for the treatment of endotoxemia in horses do exist, but the evidence for any benefit is conflicting (1).

In contrast to horses, humans most commonly receive plasma transfusions to provide coagulation factors for the prevention or treatment of bleeding disorders (4). However, even within a narrow scope of use, a lack of evidence-based medicine demonstrating any clinical benefit for such practice has generated heated discussion between protagonists and antagonists in human medicine (5). The use of plasma in equine patients has not undergone similar scrutiny. As a result, the authors of this study have questioned the clinical usefulness of transfusing plasma to hypoproteinemic horses hospitalized at our institution. In recent years, some clinicians have discontinued this practice in cases of typhlocolitis/colitis. These are the cases that typically present with hypoproteinemia. The objective of this retrospective study was to compare the clinical outcome, defined as survival to discharge, of horses with typhlocolitis/colitis that received plasma transfusion during their hospitalization at our institution, with those that did not.

The medical records of adult horses (> 1 y old) with a final diagnosis of typhlocolitis or colitis admitted to our institution between January 2000 and April 2011 were reviewed. Final diagnosis was achieved through a combination of some or all of the following: physical examination including presence of diarrhea, rectal palpation, nasogastric intubation, ultrasound examination, abdominocentesis, and necropsy. Case information retrieved for the analysis included signalment, total plasma protein (TPP) as measured by quantitative chemistry analysis, and whether plasma was administered or not. The plasma product administered was categorized as home harvest (blood collected from horses of the teaching herd), commercial, or unknown (origin of the plasma transfused not recorded). Outcome was recorded as survival to discharge from the hospital or not (euthanasia or death).

Data were statistically analyzed using a standard Fisher’s exact test to calculate the conditional maximal likelihood of the odds ratio (OR) of euthanasia/dying following not receiving plasma as opposed to receiving plasma. The exact Sterne confidence intervals (CI) on the OR were computed.

For horses that were transfused, TPP concentration (g/L) before and after (12 to 24 h) transfusion was recorded, and the change in TPP was calculated.

The total volume of plasma administered to horses that survived and to those that did not were compared using a Wilcoxon signed-rank test. To determine whether the type of plasma administered had any effect on outcome, a Chi-squared analysis was carried out to compare the use of home harvest and commercial plasma between horses that survived and those that did not survive.

A total of 465 horses met the inclusion criteria for the typhlocolitis/colitis cases during the study period. All horses had diarrhea, which was either present at the time of admission (n = 317) or became evident while in hospital (n = 148). There were 238 male and 227 female horses.

The ages ranged from 1 to 23.5 y, with a median of 12.1 y. The most common breeds represented were Thoroughbreds (n = 149), Standardbreds (n = 117), Quarter Horses (n = 47), mixed breed (n = 39), Warmblood breeds (n = 12), ponies (n = 10), Miniature horses (n = 7), and the remaining horses (n = 84), which included other breeds such as Friesian, Belgian, Haflinger, Andalusian, Tennessee Walker, and Hanoverian.

A total of 189 horses with typhlocolitis/colitis received plasma transfusion while 276 did not. Out of 189 transfused horses, 118 (62%) survived to discharge and 71 (38%) died or were euthanized during hospitalization. Out of 276 non-transfused horses, 206 (75%) survived and 70 (25%) died or were euthanatized during hospitalization. Horses with typhlocolitis/colitis that received plasma transfusion had higher odds (OR: 1.77; 95% CI: 1.17 to 2.65; P = 0.05) of dying, or euthanasia, than did non-transfused horses.

The volume of plasma transfused ranged from 1 to 25.4 L (median: 4 L), in the group of horses that survived, and 1.2 to 16 L in the non-surviving group (median: 4 L). Of the horses that were administered plasma and survived, 83 (70%) horses received home harvest plasma, 31 (26%) received commercial plasma, and in 4 (3.4%) horses the type of plasma was not recorded. In the group of horses that did not survive, 50 (70%) received home harvest plasma, 20 (28%) received commercial plasma, and in 1 (1.4%) horse the type of plasma administered was not recorded. There were no significant differences in the volume (P = 0.68) or type (P = 0.82) of plasma administered in horses that survived compared with those that did not.

In total, there were 6 (3%) mild adverse reactions at the beginning of the first transfusion. Clinical signs of a reaction included hives and increased respiratory rate and rectal temperature. Five horses with reactions were in the group that survived, and 1 horse was in the group that did not survive.

The TPP concentrations before and 24 h after transfusion were available for 115/189 (61%) and 64/71 (90%) surviving and non-surviving horses, respectively. The median TPP concentrations of survivors and non-survivors before transfusion were 42 g/L (range: 28 to 78 g/L) and 45 g/L (range: 15 to 86 g/L), respectively (Figure 1). The TPP of surviving horses increased significantly after transfusion (before: 42 g/L; range: 28 to 78 and after: 46 g/L; range: 30 to 74; P = 0.001). The TPP of non-surviving horses after transfusion was not significantly different compared with the TPP before transfusion (before: 45 g/L; range: 15 to 86 g/L, and after: 46 g/L; range: 18 to 80 g/L; P = 0.647).

Figure 1.

Total plasma protein concentrations of surviving (A) and non-surviving (B) diarrheic horses before and after plasma transfusion.

This study aimed to compare the clinical outcome, defined as survival to discharge, of horses with typhlocolitis/colitis that received plasma transfusion during hospitalization at our institution, with those that did not. Our results suggest that the administration of a plasma transfusion has a minimal association with survival. In fact, horses with typhlocolitis/colitis that received a plasma transfusion were 1.77 times more likely to die than horses that did not receive a plasma transfusion. The retrospective nature of the data collection means that a bias may exist whereby the more severely affected horses received plasma, and the severity of their disease made them more likely to die. However, notwithstanding this limitation, the data do not reveal a negative association with the lack of use of plasma. Moreover, based on recent publications, it can be argued that patients with a low TPP are not necessarily the ones most severely affected, because TPP alone has been shown to lack any prognostic value in horses presenting with acute colitis (6).

The median TPP of surviving and non-surviving horses before transfusion was consistent with hypoproteinemia. However, the range of TPP values before transfusion was wide in both groups of horses, which raises the question of whether some of the horses received plasma for reasons other than hypoproteinemia (e.g., coagulation, nutritional support, or to provide anti-endotoxin antibodies). The question of indication for plasma administration is fundamental and a cornerstone of this study. As mentioned, the rationale for the use of plasma is not evidence-based. The precise reason as to why it was administered was frequently not stated in the medical record, and therefore could not be included in the data analysis.

In the opinion of the authors, administration of plasma for coagulation was unlikely to be the rationale for treatment in this study: the efficacy of plasma transfusion in horses with identified, or suspected, coagulation abnormalities has not been proven in clinical experimental settings. Furthermore, the authors believe that the provision of anti-endotoxin antibodies was unlikely to be the reason for administration of plasma: most horses in the study received the home harvest plasma, which was obtained from horses not immunized to produce hyperimmune plasma. The practice of producing home harvest plasma was discontinued part way through the period from which the cases in this study were drawn, most likely explaining the use of commercial plasma in the remaining horses. Our results reveal that neither the type nor volume of plasma administered had any association with outcome.

Our results demonstrated a minimal increase (4 g/L) in TPP following transfusion in surviving (n = 115) horses, and a reduction (−1.5 g/L) in non-surviving (n = 64) horses. While it is tempting to make a statistical comparison of TPP among horses that survived and horses that did not survive, before and after plasma transfusion, the necessary information on the hydration status of each horse was not available through the records, rendering such comparison insignificant.

Nevertheless, the minimal increase in TPP, or lack thereof, in non-surviving and surviving horses is an interesting finding because it is comparable to findings in horses with hypoalbuminemia, in which the total plasma levels of albumin usually only increase slightly or not at all, following plasma transfusion (7). Furthermore, even when an increase in TTP has occurred after transfusion, the TPP levels tend to decrease to pre-transfusion levels within 24 h of the transfusion (7).

This observation is not unexpected if the amount of plasma usually administered and fate of the product after transfusion are considered. For example, in a 450-kg adult horse, with severe acute diarrhea, a TPP of 36 g/L, and a serum albumin concentration of 18 g/L, the calculated total intravascular albumin content would be 810 g [blood volume of 100 mL/kg body weight (BW) × 450 kg = 45 L]. Transfusion of 5 L of commercial plasma with a TPP of 62 g/L and an albumin content of 38 g/L would contribute 190 g of albumin (5 L × 38 g/L). Assuming that the transfused proteins remain in the intravascular space, and the blood volume increases by 5 L, the albumin concentration will increase from 18 g/L to 20 g/L. However, in disease states, the normal constant exchange rate of plasma constituents, such as albumin, between the intra- and extravascular compartments (trans-capillary escape rate), is increased, lowering the albumin concentration in plasma at a faster rate (8). In albumin therapy in humans, approximately 10% of the infused albumin leaves the intravascular space within 2 h, and by 2 d as much as 75% is distributed into the extravascular space (9). This distribution process can occur more rapidly if there is disruption of endothelial integrity, as occurs in colitis, leading to a significant increase in the rate of capillary albumin leakage (10,11). It is therefore possible that the increase in TPP following administration of plasma in surviving horses in this study is a reflection of a healthier state of the cecum and colon, and importantly in these horses the TPP may have increased even without a plasma transfusion.

The association between low COP in horses, and clinical outcome is not known. In humans, a relationship between death and COP could not be established in hypoproteinemic patients, and the association between serum albumin and COP in critically ill patients was found to be insignificant (12). In horses, Bellezzo et al (10) reported a discrete increase (from 11 to 13 mmHg) in the oncotic pressure after synthetic colloid administration in hypoproteinemic horses suffering from naturally occurring gastrointestinal disease. Atherton et al (11) treated colitis in horses with hyperimmunized plasma, control (regular) plasma, or no plasma. They reported that the mean values for duration of diarrhea (± standard error) were 41 ± 10 h, 119 ± 56 h, and 72 ± 25 h for the hyperimmune plasma, normal plasma, and control groups, respectively, but none of the treatments influenced the overall survival rate. Diarrhea resolved in 70% to 90% of the horses within 72 h regardless of the treatment group. The use of plasma as a source of colloidal support in horses, therefore, has a questionable impact. Large volumes may be required if plasma is used as a source of albumin. For example, approximately 25 mL/kg BW (12 L per 500 kg BW) is required to increase serum albumin by 5 g/L. It may be difficult, therefore, to justify its use with seemingly little evidence of an impact, the potential limitations of donor availability, and the economic constraints with such a large volume required.

In recent years, horses that were presented to our institution with conditions such as colitis, with TPP levels as low as 20 g/L [reference range (RR): 57 to 75 g/L], and albumin levels as low as 7 g/L (RR: 30 to 37 g/L) have been managed by addressing their primary disease and providing supportive care, without including plasma transfusions. Clinical signs traditionally associated with severe hypoproteinemia such as edema have not been observed or were only mild and transient, and considered clinically not significant. However, the change in clinical decision-making regarding the use of plasma transfusion by the authors of this study requires scientific scrutiny and further interventional studies are necessary to validate this clinical rational.

In summary, the results of this study reveal that the administration of plasma to horses with typhlocolitis/colitis did not improve the survival rate when compared to horses that did not receive a plasma transfusion. The lack of any negative association of forgoing the administration of plasma highlights the need for inquiry of current plasma transfusion recommendations, and for targeted research. The need for evidence-based medicine to support plasma transfusion is irrefutable. To provide high quality evidence supporting whether plasma has a positive effect on outcome in horses with typhlocolitis/colitis, a randomized blinded prospective multicenter clinical study would need to be conducted (13).