Neil Blumberg

There are no clinical trial data of any use yet. All the conclusions to date are based upon in vitro (laboratory) testing which has never been shown to correlate with clinical outcomes (stopping bleeding, thrombosis, etc.).

No data on clinical outcomes being equivalent or better yet, so most are not implementing because maintaining two inventories is going to be a big ask. Awaiting clinical data. Randomized trial in cardiac surgery being planned in USA.

Thanks for your generous comments Mabel. "The trauma surgeons like the idea of having some clotting factors given first (although liquid plasma on day 26 will have low levels of several labile factors)." In general I think we have overestimated the need for clotting factors and underestimated the risks of giving plasma to trauma patients, so I would disagree. Most of our level I trauma patients are hypercoagulable on TEG, not hypocoagulable. A marginally elevated PT or aPTT is not a coagulopathy and does not account for bleeding. Plasma, like red cells and platelets is likely pro-inflammatory and pro-thrombotic, although there are almost no data, unlike for red cells and platelets, where it is clear they are pro-inflammatory and prothrombotic. So we do not begin 1:1:1 MTP until after 8 units of red cells or we receive a call from the surgeon saying this patient is clinically hypocoagulable (the bleeding is diffuse and not proportional to the injury in their clinical judgement). I realize this goes against the expert opinion, but expert opinion has ignored the toxicity of plasma, platelets and possibly cryo. Obviously red cells are needed for massive hemorrhage, but I am less convinced that we are doing patients any good by starting plasma, platelets and cryo early. I think we are likely doing more harm than good. That's an advantage of low titer whole blood group O. We're not giving them unnecessary, ABO mismatched soluble and platelet cellular antigen with mismatched plasma (group AB) and platelets (group A usually). "Are you thawing plasma and keeping it for more than 5 days as "liquid plasma"? No, we are not, but we probably should be. No reason not to. Only factor VIII is lacking and these patients do not need factor VIII as their factor VIII levels are usually very high due to the trauma effect on release of factor VIII from endothelial cells. "Another question: irradiating liquid plasma won't necessarily mitigate the WBC issues that you describe, will it? The lymphocytes can't multiply but the WBC antigens, biologic mediators etc. would still be present to wreak the havoc you described, wouldn't they (depending somewhat on leukoreduction, of course)?" That's correct. There is no evidence that irradiating plasma mitigates against issues of immune modulation due to allogeneic transfusion. The reasoning for leukoreduction to my way of thinking is avoiding allogeneic white cells which cause derangements in both innate (neutrophil, monocyte, macrophage and NK cell function) and adaptive (T and B cells) immunity. Not for the better in that host defenses against infection are impaired after allogeneic transfusion. White cells also fall apart during storage so we are infusing histones and DNA which in animal models cause lung injury. When we went to universal leukoreduction our rate of TRALI went down by 83% and the rate of TACO went down 50% and stayed that way for 20 years. So leukoreduction has multiple benefits and no downside clinically. The use of non-leukoreduced whole blood is a mistake in my mind although there are no data to support it or argue against it. White cells from allogeneic donors only rarely benefit patients (decreased rejection of renal transplants is the only substantive benefit).

One exception may be New York. They require you to take an exam regardless of whether you have already passed the registry exam, if I am not mistaken. Needless bureaucracy and an absurd imposition on individual's rights to practice their profession. In the face of a real shortage of medical technologists both statewide and nationally.

Glad you mentioned washing with Plasmalyte A, which we have shown leads to less hemolysis than normal saline during 24 hours of storage. A bit more expensive, but more physiologic and hemolysis is not a good thing for patients. It is also FDA approved for use in blood component administration. When patients with acute myeloid leukemia (younger than 50, without high risk cytogenetics) receive washed platelets and red cells during induction therapy, consolidation and transplant, their mortality rate drops from 50% at five years to 20% in our hands. Also abrogates essentially all febrile and allergic reactions to transfusion, so we use it in patients with repeated or severe reactions to leukoreduced red cells and platelets. We have also never had a reported case of TRALI or TACO to about 100,000 transfused components over the last 20 years or so. The contempt for washed transfusions amongst our colleagues is both tragic and misplaced. Washed transfusions can save lives and reduce patient suffering. Also reduces inflammation post-transfusion in pediatric cardiac surgery, and may reduce mortality. Randomized trial data, by the way.

Sounds like the source of CMV in your intensive care unit likely were visitors (probably family members) or staff with new CMV infections or reactivations of previous CMV infections. The chance that there would be two transfusion transmissions in the same unit is extraordinarily low. As you say, the triumph of belief over data/science. CMV is ubiquitous in our society which is why as many as 70-90% of us are CMV seropositive. CMV testing is probably largely useless at preventing infection when transfusions are leukoreduced. I think most of the early literature may have been confounded by community acquired/person to person transmission. Just wasn't considered at the time.

Just for the record, if the blood is leukoreduced, CMV testing is redundant and adds no benefit. One less thing to complicate life. We haven't used CMV seronegative blood for any patient in 20 years and have yet to have a case of CMV associated with transfusion after >2,000 stem cell transplants and greater than 1,000 transfused premature newborns. Passive reporting, obviously, but this experience is supported by a fair amount of randomized trial and observational data. We also use recently (as in within a few days) irradiated washed red cells <21 days in storage for our newborn intensive care unit. There is no evidence that red cells any shorter in storage provide any clinical benefit. Indeed, shorter storage red cells are associated with increased nosocomial infection in randomized trials.

We convert plasma frozen within 24 hours (our standard plasma product) to liquid plasma for standby and keep some A and O units thawed at all times. For routine orders, we just thaw units as needed. The freezing and thawing of what becomes "liquid plasma" in our setting means we don't worry about graft versus host disease. By the way AB plasma is not universal donor as it is associated with worse clinical outcomes in a variety of observational studies. Lots of incompatible soluble antigen that reacts with anti-A and anti-B in 95% of recipients, injuring platelets, red cells and endothelium. We only use ABO identical plasma. Just another approach.

The mechanisms of what have been termed TRALI (actually a subset of acute lung injury/acute respiratory distress syndrome) and TACO (actually something very common, congestive heart failure) have been widely misunderstood due to unjustified assumptions/dogma. There are many biologic mediators other than antibodies that can cause lung injury after venous infusion which directly subjects the lung vascular endothelium to these mediators (antibodies, activated cells, lipids, mediators such as sCD40L, DNA/histones). Likewise there are many mediators that can cause or exacerbate cardiac failure after venous infusion (inflammatory mediators, excess volume). Cardiac failure is not just volume overload, but can be caused by fever, inflammatory cytokines and vascular/myocardial muscle dysfunction. The notion that these are distinct entities is also at variance with clinical experience. Many patients have signs of both cardiac failure and pulmonary failure simultaneously. So the definitions and pathophysiology used in reviews and texts are lacking in validity and just plain oversimplified and wrong, in my view. There are compelling data to support these iconoclastic contentions for TRALI, and some for TACO. Most germane (see attachment), when we introduced universal leukoreduction, we saw a sustained 83% drop in reports of TRALI and 50% in TACO over the following years. This suggests that white cells/DNA/histones play a role in causing lung and heart inflammation and dysfunction. This clinical observation was confirmed in animal studies from Denisa Wagner's lab at Harvard demonstrating that neutrophil extracellular traps (NETS) infused intravenously can cause acute lung injury (see attachment). To me these observations are convincing evidence that leukoreduction alters cardiorespiratory injury and failure post-transfusion and represents one of the strongest arguments for universal leukoreduction. Needless to say, this challenge to dogma has been ignored by the transfusion medicine community which continues, at least in the USA, to infuse deadly white cells and their degradation products (free DNA/histones) to patients, one of the great tragedies of the last 20 years in the USA blood bank field. We got this entirely wrong and tens of thousands of patients have probably died unnecessarily due to complications of non-leukoreduced transfusions. ULR TRALI TACO PMC version.pdf NETS and TRALI Wagner 2012.pdf

Once you've called a critical value of something that doesn't change rapidly, like a very low platelet, red cell or white cell count, you don't need to keep calling every subsequent value, since the ordering physician/NP/PA has the responsibility of checking lab values for studies they order. Not the lab. Plus it's a waste of everyone's time and annoying :). Not a big fan of critical values in general, particularly for non-emergent metrics like cell counts, creatinine, BUN, etc. Usually the clinical situation is vastly more important than any laboratory number, including for the CBC and chemistries other than electrolytes. One of the big time wasters in laboratory and clinical medicine. Only unexpected values that need action within minutes to hours need to be "critical" and called in my view. Blood cell counts aren't among them in my view. There's nothing critical about a white cell count of 500 or 100. Presence of blasts is another story.

If you are not irradiating red cells, irradiating plasma makes little sense, since there are likely many more residual leukocytes in red cells that have been leukoreduced than in liquid plasma. I don't have exact data so you may want to confirm that. I don't know the methods currently in use, but the ideal would be to leukoreduce whole blood so the plasma made afterward would have many fewer white cells. White cells are bad for patients, causing platelet refractoriness, transfusion reactions, immunomodulation predisposing to sepsis/infection and probably TRALI and TACO. So we want to transfuse as few white cells as possible, regardless of irradiation. Why are you using liquid instead of frozen stored plasma, out of curiosity? Price? The factor VIII/vwF content is likely quite a bit lower, which may not be a big deal, but isn't likely an advantage.

In our institution the most inappropriate use of group O cells is for trauma patients where we haven't received a sample so that we can switch patients to their own ABO type. We exhort our trauma team to send the first sample they send to us. ABO mismatched red cells is a suboptimal use of group O red cells, and in the long run, probably harms patients. We switch patients as soon as we can when they are A, B or AB to their own red cell type. I know there are some old dogmas that you should keep patients on group O once you've started. No data whatever to support that, practice some data suggesting it's harmful and it's obviously terrible for the supply of group O red cells.

"Do you perform all the antigen typing on your cord samples so you know which ones to use?!" Yes, we now actually use a four cell panel of frozen cord cells that last about a month in Alsever's solution once thawed. We select cells so we have appropriate negatives and positives for the most important clinically significant antigens. We don't concern ourselves with Lewis, P, etc. We do a full phenotype on the selected cells we use. Labor intensive but otherwise inexpensive.

Those are good points Malcom. We have yet to see any new alloimmunizations after >100 patients tested, probably because these patients have dramatically decreased B cell function. Fortunately few of these patients need red cell transfusions as most of the new treatments for multiple myeloma, and there are many, are not causes of severe anemia.

If you have access to cord red cells from your OB service, these are negative for CD38, and we use a panel of three of them to rule out alloantibodies when patients are receiving daratumumab(Darzalex). Transfusion . 2015 Sep;55(9):2292-3. doi: 10.1111/trf.13174.

Unless the pharmacy is going to pay attention to ABO and Rh, I would think this belongs in the transfusion service. Their computer system probably isn't set up to worry about such things :). And please do not give AB plasma to everyone. This is one of the oldest but most dangerous concepts in transfusion medicine. AB plasma is associated with increased bleeding, sepsis, lung injury and mortality when given to group O patients, for example. ABO identical is by far the most effective and safest when plasma must be transfused. Happy to provide references. There is no universal plasma, contrary to dogma and long practice.

Malcolm, I agree this was truly insensitive and arrogant behavior, not atypical of physicians in many periods of history, granted. I do believe this is very different from our use of about to be discarded cord blood specimens used for laboratory testing, but I'm sure some would disagree. Our human subjects review boards generally consider laboratory testing blood specimens from patients that would otherwise be discarded may be used for research purposes or quality control purposes without informed consent.

We have yet to see a patient on daratumamab who has made an anti-K antibody after years of transfusing red cells without regard to K antigen status. We use cord red cells in an antibody screen to rule out significant antibodies to allogeneic red cells (they are CD38 negative) as our method of dealing with this issue of pan-reactivity from daratumumab. I know this practice isn't allowed in the UK due to the over the top regulations that followed the infant parts kerfuffle. We detect plenty of anti-K's, just NOT in patients receiving anti-B cell therapies. In fact, I cannot recall a single new red cell alloantibody in myeloma or lymphoma patients receiving daratumumaub, rituximab, etc. No B cells equals dramatically reduced risk of alloimmunization, so you may be worrying about something that is pretty unlikely to happen. Just another approach.

Blood ABO major incompatible platelets increase bleeding in intracranial hemorrhage . 2021 May 13;137(19):2699-2703. doi: 10.1182/blood.2020008381.

I think the evidence (mostly adults) is that AB plasma is no safer than A plasma, probably because it is soluble antigen incompatible with 95% of patients compared with 60% for A plasma. Obviously ABO identical is safest and probably most effective. With platelets, giving ABO major incompatible platelets increases bleeding in adults so I would definitely not be preferring AB platelets which are incompatible with 95% of recipients. At this point, group O may be safest and most effective for initial transfusions until ABO identical can be transfused. We've worried about hemolysis but have neglected to detect that transfusing incompatible antigen (A and B cellular and soluble antigen) forms immune complexes that contribute to hemolysis even when the incompatible antigen isn't from red cells. This is a rapidly evolving field with not a lot of clinical outcome data, much less randomized trials. Nonetheless, it appears we've gotten the ABO system wrong when it comes to what is safest. And it isn't AB plasma, which, in adults, is associated with with increased sepsis, lung injury and mortality in group O patients. There are no data, but I assume the same would be true for AB platelets given to O recipients. So our choice would be O red cells (or low titer group O whole blood), A plasma and A or O platelets. We would never use AB plasma or platelets except for AB patients, given the bad outcomes in adult patients. Arch Surg . 2010 Sep;145(9):899-906. doi: 10.1001/archsurg.2010.175. Vox Sang . 2009 May;96(4):316-23. doi: 10.1111/j.1423-0410.2009.01167.x. Epub 2009 Feb 24.

I think the evidence (mostly adults) is that AB plasma is no safer than A plasma, probably because it is soluble antigen incompatible with 95% of patients compared with 60% for A plasma. Obviously ABO identical is safest and probably most effective. With platelets, giving ABO major incompatible platelets increases bleeding in adults so I would definitely not be preferring AB platelets which are incompatible with 95% of recipients. At this point, group O may be safest and most effective for initial transfusions until ABO identical can be transfused. We've worried about hemolysis but have neglected to detect that transfusing incompatible antigen (A and B cellular and soluble antigen) forms immune complexes that contribute to hemolysis even when the incompatible antigen isn't from red cells. This is a rapidly evolving field with not a lot of clinical outcome data, much less randomized trials. Nonetheless, it appears we've gotten the ABO system wrong when it comes to what is safest. And it isn't AB plasma, which, in adults, is associated with with increased sepsis, lung injury and mortality in group O patients. Arch Surg . 2010 Sep;145(9):899-906. doi: 10.1001/archsurg.2010.175. Sang . 2009 May;96(4):316-23. doi: 10.1111/j.1423-0410.2009.01167.x. Epub 2009 Feb 24.

Disciplinary action over product wastage sounds like it is perhaps well intentioned but ignorant bureaucrats, not health care workers running the show. That's a big part of the problem in many institutions these days. We are fortunate in that the senior decision makers in our hospital are all physicians, nurses, etc., including the CEO, CMO, COO. Washing your hands before delivering babies turns out to be inconvenient but a better idea. Universal leukoreduction and avoiding infusion of ABO incompatible antigen and antibody are also better ideas than what we have done for decades or longer. These practices will save lives, reduce need for transfusions and actually save the system money overall, albeit at greater expense in the transfusion service. You are no doubt correct that there will be pushback from transfusion service staff used to doing things the old, easy, but harmful way, and hospital administrators who prioritize the wrong things such as budgetary tunnel vision over reduced harm. And blood centers are not likely to initially be all that interested in changing practices. But when the data says patients do better with universal leukoreduction and ABO matching, hopefully, in the long run the dogma will be replaced by data driven practices. May take a while.

Once clinicians understand that ABO mismatched platelets not only do not provide hemostasis, but make bleeding more likely/worse, they will be less willing to accept infusion of ABO mismatched antibody/antigen. Once blood transfusion services realize that infusing ABO mismatched platelets increases utilization by two fold, they will be more interested in making the effort to give ABO identical or remove incompatible plasma by washing. Doing the right thing for patients is never the wrong answer to the question. Our current practices are convenient for us and minimize waste. We need to prioritize clinical benefit over inventory control and waste reduction. What we are doing now is providing little to no benefit and actually harming patients in many instances. The bleeding rate in the PLADO (platelet dose study) in NEJM was 70%. That's not exactly a clinical triumph. Our bleeding rate is probably less than 5-10% employing ABO identical/washed platelets.

Cannot post the entire article due to copyright restrictions, but most institutions have access to NEJM through their library. If not, shoot me an email at neil_blumberg@urmc.rochester.edu and I'll send along the .pdf. If you are transfusing 40-60 platelets a day, giving ABO identical to group O and A individuals should be relatively easy. When patients are changing ABO blood group it becomes more difficult. We avoid transfusion ABO antigen and/or antibody that is incompatible with either original recipient type or donor type. Usually means washed group O red cells and platelets. That's the bad news. It does require time and effort, and as you say, med techs are in short supply. Here's the good news. If you transfuse ABO identical or washed compatible platelets you will use between 30-50% fewer platelets per patient, increasing your supply and decreasing your cost/problems. You will also use next to no HLA matched platelets (we used 3 out of 6,000 one recent year), you will have fewer febrile and allergic transfusion reactions, you will have fewer red cell as well as HLA antibodies made in recipients, and you may reduce TRALI and TACO. Obviously you have to have universal leukoreduction to start with. Selective leukoreduction misses about 50% of the patients who will become refractory, probably due to missed or delayed diagnosis of hematologic malignancy, aplastic anemia, etc. But the big attraction is you will have less bleeding, although that mainly affects the patients and the docs and nurses at the bedside. When you transfuse ABO major incompatible, which seems to be the default due to fear of hemolysis from minor incompatible, you don't get any increments, you use lots of platelets and the patients bleed more. (see references below) Bleeding causes lots of harm, but also impacts the blood transfusion service for obvious reasons. So figure out a way to start giving patients with aplastic anemia and acute myeloid leukemia who are newly diagnosed only ABO identical platelets and that will be a great start for the patients and the transfusion service. Those patients will bleed less, need fewer platelet transfusions, have fewer transfusion reactions, will not have positive DATs, and will likely survive their hospitalization and disease at higher rates if our experience is typical. And if you cannot give ABO identical or washed platelets free of incompatible cellular and soluble antigen and free of incompatible ABO antibody, start out with minor incompatible platelets (e.g., O to A) rather than ABO major incompatible (e.g., A to 0). The risks of hemolysis are not negligible (about 1 in 800) but are less serious and severe than having life threatening bleeding or refractoriness which occur more rapidly with ABO major incompatible in all likelihood. There's a ton of antibody that is incompatible with antigen transfused when we give A platelets to O recipients which means each antigen winds up with a ton of antibody making huge immune complexes. When we transfuse antibody incompatible we are transfusing a small amount of antibody into a recipient with huge amounts of antigen, so the size and number of immune complexes is probably smaller. These are my best guesses that we've been making exactly the wrong decision when we give ABO mismatched platelets. Best to avoid any, but major mismatched provides no hemostasis, minimal to no increment and is associated with increased bleeding mortality in the study from Columbia (David Roh and colleagues https://pubmed.ncbi.nlm.nih.gov/33649761/). But ABO identical is not that hard for larger centers for the 85% of patients who are group O or A. You just have to start small, get the hang of it, and then extend to other blood groups and other diseases than leukemia, MDS and aplastic anemia (including transplants, particularly allo--transplants). All those tables of how to select ABO mismatched platelets for transplant recipients are well intentioned but scientifically without evidence. Avoid infusing incompatible antigen and antibody as much as possible, and delay transfusion when ABO identical will be available within hours. Give priority to patient well being over inventory management. Give reduced doses, which work just as well. Get a Terumo or Haemonetics washing device and wash with PAS. It's a big set of changes, but neither terribly expensive nor rocket science. The dogmas and expert opinion about universal leukoreduction and ABO matching of transfusions are now proven to be tragically mistaken. https://www.ashclinicalnews.org/news/from-the-blood-journals/written-in-blood/outcomes-abo-incompatible-platelet-transfusions-patients-intracerebral-hemorrhage/ https://pubmed.ncbi.nlm.nih.gov/11399821/ https://pubmed.ncbi.nlm.nih.gov/21414009/

Research letter in NEJM describes our findings. https://www.nejm.org/doi/full/10.1056/NEJMc2034764?fbclid=IwAR1BQRvpaHBAMDaHxCPY07xBjPQHlIHoJCOmpjoT_pBNvQsV7pzzDVdLYaY Table 1. HLA-Matched Platelets as a Percentage of All Platelet Transfusions, According to the Initiation of Other Protocols.* Protocol and Timing of Initiation No. of Years HLA-Matched Platelets Difference from Previous Period (95% CI) median % (IQR) percentage points No leukoreduction and no ABO matching (1985–1990) 6 12.5 (4.2–13.7) NA Leukoreduction and ABO matching for patients with leukemia and MDS (1991–1999)† 5 2.9 (2.0–3.6) 9.6 (9.4–9.8) Universal leukoreduction (2001–2004) 4 1.4 (1.1–2.0) 1.5 (1.4–1.6) Universal ABO matching (2005–2015) 11 0.4 (0.2–0.8) 1.0 (1.0–1.1) Pathogen reduction of platelets (2016–2019)‡ 4 1.7 (0.8–1.8) ND The practical fact is that this can only be implemented by medical technologists, not physicians, and this is a daunting prospect. But the reality is that ABO mismatched platelet transfusions probably exacerbate rather than prevent bleeding, so waiting for ABO identical is likely better than just transfusing whatever is available. This is going to require a major change of approach because the (incorrect) dogma, based upon no data, is that ABO doesn't matter for platelets. I understand why many people's instant reaction is this is not feasible. But it is once technical experts in your transfusion service figure out how to implement it gradually. In our randomized trial back in 1993 (see ref below), the ABO identical group only required <50% the number of platelet transfusions (every other day instead of every day on average) compared with the usual first in, first out regardless of ABO type group. ABO mismatched transfusions create a hostile environment with gigantic immune complexes that compromise subsequent transfusion that may be ABO identical. Avoiding this is key. Our suggestion is to start gradually. Pick new previously untransfused patients with aplastic anemia and good prognosis AML (young, favorable or normal cytogenetics), groups O or A, and start with them. Eminently doable since your platelet supply is 45% O and 40% A, just like your patients, on average. Get the hang of doing this and prioritizing ABO for at least some patients. Once you get this in place, extending it to other patients and eventually all patients can happen. We use washed O or A platelets and red cells for group B and AB recipients when we don't have group B or AB platelets. No increase in bleeding and fewer transfusion reactions, lung injury and congestive heart failure (what we call TRALI and TACO). In the end, you have more surviving patients and fewer headaches and transfuse many fewer platelets per patient and essentially no HLA if you can get your referring hospitals to stop our current standard harmful practices. Godspeed. Heal JM, Rowe JM, McMican A, Masel D, Finke C, Blumberg N. The role of ABO matching in platelet transfusion. Eur J Haematol. 1993 Feb;50(2):110-7. doi: 10.1111/j.1600-0609.1993.tb00150.x. PMID: 8440356. Happy to have discussions or visitors so our technical experts can show you how it is feasible.