Evidence-Based Plasma Transfusion Carolyn Burns MD, Physician Advisor

Evidence-based Plasma Transfusion Carolyn Burns MD, physician advisor

Current guidelines include the following indications for plasma Plasma is prepared by transfusion: centrifugation from whole blood donation. The familiar standard is  Management of patients with coagulopathy, congenital or acquired, for which FFP (fresh frozen plasma), which specific factors concentrates are not available is, by regulation, plasma frozen  Management of bleeding patients who necessitate multiple factors, including within 8 hours of whole blood massive hemorrhage protocols donation (WBD). Plasma 24 is  As replacement fluid for patients undergoing plasma exchange for treatment of plasma frozen within 24 hours of thrombotic thrombocytopenic purpura (TTP) WBD. This allows more time for  Management of patients with rare C1-esterase deficiency for whom specific donor centers to process whole blood, opening a broader factor concentrates or drugs are not available timeframe and donor outreach.

Thawed plasma has been of benefit to hospital transfusion services as this component represents FFP or Plasma 24 which has been thawed, not utilized, and returned to the transfusion service. This product may be relabeled and stored for five days at 1 to 6 degrees Celsius. Prior to the approval of this product, plasma that was thawed and returned had an expiration of 24 hours, often resulting in significant wastage. Thawed plasma is most beneficial for hospitals with high volume plasma use, e.g. trauma centers, high-risk cardiovascular surgery and obstetrics, and solid organ transplant. It is an excellent way to maximize inventory, prevent waste, and provide rapid access. All three of these plasma products have relatively equivalent content of coagulation factors. There are additional plasma components, such as liquid plasma, cryo-poor plasma, and solvent/detergent plasma, but these are not frequently utilized and thus we will not include these in routine discussion.

The content of plasma represents blood proteins, i.e. albumin, fibrinogen,

procoagulant and anticoagulant factors, as well as immunoglobulins and antibodies such as anti A & B; thus, plasma must be ABO compatible. Each unit is

roughly 250 to 300 mls. The typical adult dose (if truly indicated) is roughly 4 Evidence Provides Clear units based on the size of the patient. If one calculates strictly, then a dose of Contraindications for plasma should be 10-15 mls./kg BW. Strict calculation is most important in Plasma: neonates, the elderly, and those at risk for volume overload. The timing of plasma transfusion is very important as several coagulation factors, specifically factor V, Do not transfuse plasma for the factor VII, and vWF, have relatively short t1/2s of around 4-6 hours. Thus, following: providing plasma within this timeframe for a planned procedure (again, IF truly indicated) would ensure adequate plasma levels. Follow-up testing is • Volume replacement or nutrition recommended within 15 to 60 minutes post-transfusion to assess the response. • Disseminated intravascular Current guidelines include the following indications for plasma transfusion: coagulopathy without bleeding management of patients with coagulopathy, congenital or acquired, for which • Fluid replacement during plasma specific factors concentrates are not available; management of bleeding patients exchange for entities other than who necessitate multiple factors (this includes massive hemorrhage protocols), as TTP replacement fluid for patients undergoing plasma exchange for treatment of • To prevent periventricular thrombotic thrombocytopenic purpura (TTP); and management of patients with hemorrhage in premature rare C1-esterase deficiency for whom specific factor concentrates or drugs are not neonates available. Additionally, evidence provides clear contraindications for plasma. Do not transfuse plasma for the following: volume replacement or nutrition, • As prophylaxis for mild to disseminated intravascular coagulopathy without bleeding, fluid replacement moderate elevations of INR in during plasma exchange for entities other than TTP to prevent periventricular non-bleeding patients hemorrhage in premature neonates, or as prophylaxis for mild to moderate elevations of INR in non-bleeding patients. The last situation is worth considerable discussion, as this clinical scenario represents the majority of plasma use and current literature points to a lack of evidence for prophylactic plasma transfusion, especially in the peri-procedural arena.

Plasma transfusions are associated with adverse events and poor outcomes similar to what we see with other blood components. Immune-modulation, increased infections, LOS, ventilator dependence, IN A SURVEY OF PRACTICE, and increased CA recurrence have been associated in many studies with RBC transfusions. New evidence is pointing toward similar findings 72% OF PHYSICIANS ADMITTED TO associated with plasma transfusion, specifically infections and multi- NONTHERAPEUTIC ADMINISTRATION organ failure. In particular, TACO and TRALI, significant pulmonary complications, are often associated with plasma. TACO and TRALI OF FFP. remain the No. 1 and No. 2 causes of transfusion-associated-death as reported to the U.S. Food and Drug Administration. Allergic and CALDWELL S ET AL. HEPATOLOGY 2006; 44:1039 anaphylactic reactions, as well, can be serious and life-threatening. These risks must be recognized when considering transfusion. All of these potential adverse events should be considered as part of our medical decision to transfuse, discussion of patient choice, and the consent process.

So why do HCPs most often transfuse plasma outside of active hemorrhage? Interesting and a bit alarming, a 2006 study highlighted that plasma transfusions were performed for perceived medico-legal risk rather than true therapeutic goals, with 72 percent of physicians admitting to nontherapeutic administration of FFP in a survey of practice.1 Therefore, it is critical that we all acknowledge and embrace the most recent evidence. Plasma, as with RBC and platelet transfusion, should not be transfused without a clear indication.

With the serious, potentially life-threatening adverse events associated with plasma in mind, we should focus on the most frequent use of plasma, i.e. prophylactic transfusion in the non-bleeding patient. Let’s explore the reasons why HCPs might feel compelled to give plasma, and then we will launch into the literature that provides contrary evidence. Again, we will not be discussing the exsanguinating patient as everyone would agree that plasma will be included in hemorrhage protocols. Let’s begin with three assumptions as outlined in Dr. Paul Mintz’s textbook on transfusion practice:2 1) abnormal coagulation studies can accurately predict bleeding risk, 2) transfusion of plasma prior to procedures will correct the presumed hemostatic abnormality, and 3) prophylactic transfusion given prior to a procedure is of greater benefit than therapeutic transfusion given after the procedure if clinical bleeding actually occurs. These assumptions are misplaced and will lead to inappropriate and unnecessary plasma transfusion. What are some basics about plasma that must be remembered?

Why might plasma not be effective when used prophylactically? First, all plasma is not “created equal.” Plasma derived from WBD has a similar volume, as mentioned; however, there is not an equal amount of plasma proteins. Each of us has different quantities and activity levels of our procoagulant factors. Even different blood types are associated with different levels. Donor centers do not quantify the factor levels in each bag of plasma. This is not required.

An interesting study by Holland et al. showed that INRs performed on blood donors showed a range of values with normal healthy donors having INRs outside the reference range.3 Does this imply that the person with an INR of 1.4 or 1.5 that is otherwise healthy is prone to bleeding? Indeed not. A subsequent study by the same investigators also showed that transfusing plasma to those with mildly elevated INRs (<1.7) did not result in a significant decrease in the post-transfusion INR.4 The INR does not have a linear relationship to coagulation factor levels, as noted in this accompanying graph. A 2018 study supports this.5 In more than 6,700 non-bleeding patients with elevated INR at a single institution, 20 percent, or one-fifth, received prophylactic plasma. A “normal” INR was only achieved in 12 percent of patients. To drive this point further, what is a normal INR? In actuality, the INR was developed as a mathematical calculation to standardize the PT for patients taking a Vitamin K antagonist (VKA), such as warfarin. The normal reference range provided by the laboratory represents a sampling of the demographic, resulting in a typical bell-shaped Gaussian curve, which thus implies some lab values outside of two standard deviations. As noted in the original Holland study, the INR does not reflect bleeding risk and should reflect a therapeutic range more than a so- called normal range.

It was not until the latter part of 2010 that evidence-based guidelines were reviewed using the GRADE methodology to look at clinical situations where plasma was typically transfused. Roback and colleagues found only two scenarios that showed the necessity for plasma transfusion.6 These were massive transfusion in trauma and active intracranial hemorrhage in patients on VKAs. This latter situation was prior to the approval of 4-factor prothrombin complex concentrates (4FPCC). Plasma is no longer the intervention of choice, as noted earlier. No clinical scenarios without active bleeding were found to meet the GRADE criteria for transfusion of plasma. The subsequent meta-analysis by Yang et al. in 2012 found no benefit, either prophylactic or therapeutic, in any of the listed clinical arena.7 Trauma hemorrhage was not included.

McCully et al. published a study of stable trauma and surgical patients and found that although the INR “normalized” after plasma transfusion in numerous patients, there were no differences in other coagulation studies, including TEG or specific clotting factor assays.8 They concurred that an isolated abnormal INR in the non-bleeding patient does not reflect coagulopathy and plasma should be limited to bleeding patients.

A study from Triulzi et al. provides a snapshot of the epidemiology of plasma use in the United States.9 It highlights that the majority of plasma is not used for bleeding patients in our emergency departments or surgical suites, but rather in general medical/surgical and intensive care units in non-bleeding patients; in fact, 80 percent. Note that more than half of these prophylactic transfusions were provided as one to two units, which is well below the recommended adult dose. Furthermore, less than one-half of plasma transfusions resulted in an INR < 1.6, mirroring the findings of the Holland and Warner studies.

The TOPIC Trial was an important RCT that, although small, characterized the real-world use of prophylactic plasma transfusion.10 It was performed in the acute-care setting on non-bleeding patients who had abnormal coagulation studies, INR ranging from 1.5 to 3.0. These authors clearly showed no differences in bleeding complications in critical care patients undergoing a vast array of minimally invasive procedures, regardless of receipt of prophylactic plasma.

A cohort study of more than 1,200 non-cardiac, non-bleeding patients with INR > 1.5 also showed a lack of evidence for prophylactic plasma transfusion.11 This study found actual increases in WHO grade bleeding episodes in those who received prophylactic plasma. The authors concluded that current evidence suggests, again, conservative use of plasma in non-bleeding patients.

A study from the Mayo Clinic reviewed outcomes in more than 18,000 patients, of whom almost 10 percent received prophylactic plasma transfusion prior to interventional radiologic procedures.12 Plasma transfusion was associated with increased need for RBC transfusion, a surrogate marker for bleeding, and admission to the ICU post-procedure. Thus, plasma can cause harm to patients and does not necessarily provide better outcomes.

Specific discussion for patients needing neuro-axis procedures should have its due. Studies show normal hemostatic levels in neurosurgical patients with intracranial hemorrhage when INR is less than or equal to 1.7, and 55 percent of bleeding episodes in neurosurgical patients occurred in those with normal pre-operative coagulation studies.13,14 The most recent American Society of Anesthesiologists guidance recommends plasma only if bleeding and INR ≥ 2.0.15

Evidence-based guidelines for plasma transfusions in patients with end-stage liver disease (ESLD) in need of gastrointestinal interventions have been recently published.16 In the past, most clinicians would attempt to have the INR lowered to 1.5 to 1.6 utilizing plasma. Current guidelines do not recommend plasma in ESLD patients with an INR is ≤ 2.5. If ≥ 2.6, Vitamin K and plasma may be considered. Patients with ESLD have a rebalanced hemostasis with decline in procoagulant AND anticoagulant factors. They have normal thrombin generation and are often more likely to clot than bleed.

The most recent 2018 monograph from the British Society of Hematology incorporates findings from these studies as well as a multitude of others.17 The society clearly states that the use of prophylactic plasma prior to procedure in non- bleeding patients with abnormal coagulation studies is not supported by good quality evidence, and these studies do not predict peri-procedural bleeding.

Plasma is no longer recommended for emergent reversal of VKA-associated bleeding given the approval of 4FPCC. The real-world experience with IV The impact of commonly utilized Vitamin K and 4FPCC eliminates the need for plasma in this patient population. doses of plasma to correct Studies using this approach showed shorter times to achieve hemostasis as abnormal coagulation studies or well as fewer AEs such as TRALI or TACO.18 In a 2016 meta-analysis, which included RCTS and observational studies, the combination of IV Vitamin K and to reduce bleeding risk is very 4FPCC for emergent warfarin reversal was associated with a significant limited, particularly when the decrease in mortality, achieved a rapid reversal of INR, and was associated with INR is between 1.5 and 1.9. fewer adverse events such as volume overload.19 There were no significant thrombo-embolic events. This approach should be incorporated into anticoagulant reversal protocols.

Patient Blood Management (PBM) promotes quality, safe and effective patient care as well as resource management and fiscal responsibility. Thus, a quick discussion regarding the cost of plasma is necessary as clinicians may consider emergent reversal of VKA and the expense of 4FPCC. We do have a responsibility to be good stewards of resources and health care dollars. Activity-based cost (ABC) analyses can provide insight. An ABC analysis published in 2010 revealed the actual cost of an RBC unit was three- to four-times greater than the direct acquisition cost from the donor center.20 An identical ABC analysis applied to plasma showed a 10-times increase from the acquisition cost per unit.21 Therefore, plasma is not an inexpensive component. In particular, if plasma is used for VKA reversal, often two to four adult doses are needed over a 24-hour time period, which potentially results in a “cost” of $2,000 to $4,000. One must also consider the potential “cost” of TRALI and/or TACO. Restrictive plasma transfusion not only provides the safer, more effective care we strive for, but also has significant fiscal and community benefits.

Several institutions have utilized this robust literature to implement change in their plasma transfusion practice, bringing restrictive transfusion of this component to the forefront. A group from the University of Texas Southwestern Medical School incorporated education and prospective monitoring of plasma to triage orders.22 Plasma use was reduced by 60 percent. The PBM team from Rhode Island has shown in two serial studies that with active education, discussion and engagement of all health care providers, significant and sustainable reduction in plasma use can be seen, encouraging appropriate use and no differences in bleeding outcomes or mortality.23,24

Given the robust and current evidence regarding plasma, prophylactic transfusions should be avoided, with very few exceptions. Careful consideration of the patient’s clinical

Figure 1-7. Zones of response to bleeding risk at the time of invasive procedures. The x-axis is meant to depict the product of both platelet number and functional activity. Patients with normal laboratory values are represented by the smallest rectangle. A large number of patients with mild-to-moderate abnormalities of pre-procedure laboratory tests are in the zone of physiologic reserve and not likely to derive any benefit from pre-procedure 1 transfusion therapy. Mintz P ed. Transfusion Therapy: Clinical Principles and Practice, 3rd ed, AABB Press, 2011. condition, aligned with this solid evidence, will prevent unnecessary and potentially harmful plasma transfusions. A nice visual is provided from Dr. Mintz’s textbook.2

PBM is the gold standard. It implies patient centricity and safety. PBM practices bring an evidence-based approach to managing anemia and minimizing bleeding and blood loss, with multimodal interventions to reduce and potentially eliminate the need for transfusion.

References

1. Caldwell S et al. Hepatology 2006; 44:1039 rd 2. Mintz P Ed. Transfusion Therapy: Clinical Principles and Practice, 3 ed, AABB Press, 2011. 3. Holland L et. al. Transfus, 2005;45:1234 4. Holland L et al. AJCP, 2006; 126:133 5. Warner M et al. Anesth Analg 2018; 127: 349 6. Roback J et al. Transfus 2010; 50:1227 7. Yang L et al. Transfus 2012; 52:1673 8. McCully S et al. J Acute Care Surg 2013; 75:947 9. Triulzi D et al. Transfus 2015; 55:1313-1319 10. Muller M et al. Transfusion 2015: 55:26 11. Jia Q et al. Lancet Haemotol 2016;3:e1393 12. Warner M et al. Mayo Clin Proc 2016; 91:1045 13. Matevosyan K et al. Neurosurg 2011; 114:3 14. Schramm et al. Anesthes Intens Care 2001; 29: 388 15. American Society Anesthesiologists Task Force Anesthesiol 2015; 122:241 16. Yates S et al. Transfus 2016; 56:791-798 17. Green L et al. Brit J Haematol 2018; 181:54 18. Goldstein J et al. Lancet 2015; 385: 2077 19. Chai-Adisaksopha C et al. Thromb Haemost; 2016; 116: 879-890 20. Shander A et al. Transfus 2010; 50: 753-765 21. Shander A et al. Vox Sang 2016; 111:55 22. Sarode R et al. Transfus 2010: 50: 487 23. Tavares M et al. Transfus 2011; 51:754 24. Nixon C et al. Transfus 2017; 57: 1863