Neil Blumberg

"the presenter stated specifically that NOTHING has been grandfathered. " I think the presenter is mistaken. The FDA specifically noted that the area of rare reagents and cells, and similar testing would not be subject to LDT enforcement. Of course, all the opinions in the world matter not a bit until the FDA actually acts or does not act. I cannot imagine they want to be inspecting every tertiary care hospital and blood center reference laboratory for this purpose. And most of the things we are discussing in the transfusion service and immunohematology lab are not used to provide diagnostic results to practitioners, but rather used for internal resolution of therapeutic decisions. Quite different from your average laboratory test which provides quantitative or semi-quantitative result to physicians and other practitioners who make decisions based upon lab results. Perhaps a nuance, but a real difference. If the FDA insists we validate the use of a potent anti-HPA1 anti-platelet antibody in our decision making, we're out of luck :). Ain't happening. Interestingly, much of what we do in clinical medicine has not been "validated" or subjected to FDA-like regulation. Such as using autologous or allogeneic stem cell transplants, liver transplants, using a stethoscope or looking at a patient's retina with an ophthalmoscope. No validation. No data to speak of at all.

I think most blood bank reagents and tests have been grandparented in. The FDA knows there is no alternative to these home brew reagents and testing procdures.

There are no data to answer your questions, as far as I know. It's important to make sure that the fetal cell quantitation is not measuring maternal cells with increased fetal hemoglobin, as this would overestimate the RhIgG dose needed. This is not a problem with some methods (anti-Rh(D) quantitation of fetal cells but can be a problem with acid elution (K-B) staining, for examples. If there is convincing evidence these are fetal cells, give the correct dose IV even if many vials. IM injections are cruel and unusual punishment if IV injectables are available. For patients who are not planning future pregnancies, this should be discussed with the patient. For sick patients who have received transfusions, we do not infuse RhIgG except for younger women (<40-50) who plan future pregnancies and have a prognosis for survival. Hemolysis from RhIgG can be a problem at high doses of RhIgG and large transfusion volumes. On balance we usually elect not to give RhIgG to women who have received entire or multiple units of Rh(D) positive red cells. It's a complex clinical decision with little science to guide us.

We do not use enzyme treated cells when trying to detect or work up cold agglutinins. No reason to enhance their reactivity in vitro. If they are not detectable by routine LISS at body temperature or antiglobulin methods they are not of clinical importance.

Another inspector who is a bureaucratic and clinically ignorant rigid thinker. My sympathies. There is no reason to document early infusion rates, and these vary by patient due to clinical condition. This is just a guideline and not a requirement, as the inspector would know if they had any bedside clinical practice experience. Just say it's a rough guideline, not a requirement, and clinical judgement will determine the infusion rate for each patient.

If the antibody does not react at 37 or antiglobulin phase, we would not recommend a blood warmer. Agree that in patients with cold agglutinin disease or hemolysis after exposure to cold, would use a warmer. Remember that our techniques in the lab routinely detect antibodies of no clinical significance. For one thing, we never centrifuge patients at 170g to enhance binding of red cells to antibody :). So on balance, unless you have an antibody of known clinical significance historically (antibodies in the Rh, Kell, Duffy, Kidd, S, etc. systems) and it reacts at body temperature, it's probably not significant. Cold antibodies of wide thermal range may be significant but usually not if they don't react at body temperature in vitro.

As you know a difficult question. The use of group O blood for non-O patients should be limited as soon as the ABO type is known, and ABO type specific given. There are some old admonitions to never transfuse type specific after X numbers of group O red cells or whole blood. This is nonsense, so do not follow this old precept. Give only ABO type specific as soon as possible. As for giving non-O blood to O patients when all other options are exhausted, I would only do this if death were imminent due to bleeding, not ever for routine transfusion. There are case reports of no hemolysis in such situations, including the one below from my original mentors from half a century ago. Accidental error but no consequences. But giving non-O blood to O recipients has the potential to cause rapid death in many instances. We don't know why there are such varied responses. Case Reports Transfusion . 1975 Nov-Dec;15(6):577-82. doi: 10.1046/j.1537-2995.1975.15676082233.x. Unusual response to ABO incompatible blood transfusion D H Buchholz, J R Bove PMID: 1198685 DOI: 10.1046/j.1537-2995.1975.15676082233.x Abstract Three units of group A blood were inadvertently administered to a group O recipient during surgery without evidence of hemoglobinemia, hemoglobinuria, hypotension, disseminated intravascular coagulation, acute renal tubular necrosis, or other signs and symptoms of transfusion reaction. The recipient had normal concentrations of IgG, IgA, and IgM as well as complement (C3) prior to transfusion and anti-A agglutinins titered to 64 (titer of 128 by the antiglobulin technic). Seventeen hours following the transfusion, 28 per cent of the circulating red blood cells were group A (equivalent to 475 ml of packed cells); they were eliminated by day 5 without evidence of hemoglobinuria, hemoglobinemia or hyperbilirubinemia. Anti-A titers (antiglobulin) had risen from a posttransfusion low of 4 to 4,096 by day 10. After treatment of serum with 2-mercaptoethanol, however, hemolytic activity which was first noted on day 5 was lost and the antiglobulin titer dropped to 24 which suggested that most of the anti-A produced in response to the transfusion was IgM rather than IgG. The anti-A titer had dropped to essentialyy pretransfusion levels and the majority of anti-A present was IgM by day 91. The recipient suffered no untoward effects from the transfusion and was in good health three months following the transfusion.

Transfusion has much more serious adverse effects than making an anti-D. Increases in infection, sepsis, thrombosis, inflammation and mortality for example. There are no data to my knowledge of long term effects of anti-D formation in patients not having future pregnancies. Most such patients come to the attention of the transfusion service because they have anti-D or simply because they are Rh (D) negative. They are then transfused with D negative blood if need be, in something like 99.99% of cases. The rare patient who gets Rh positive blood (trauma patients) do sometimes have increases in bilirubin, LDH, etc. and delayed or rarely acute transfusion reactions. These are bad for patients, so you are right, for these rare patients, the outcomes can be dire. But there few alternatives to transfusing Rh (D) positive blood to most patients in emergencies. And very few will have future transfusion reactions.

Depends on the distance from transfusion service to ED/OR. I assume you are not planning on having the cooler or refrigerator with blood in the ED on a routine basis, only when the mass casualty event occurs? Are you the only hospital in the area, or are there other hospitals with level I or II trauma capability?

It varies from no reaction to lethal hemolysis. Anti-D is not entirely predictable in causing severe hemolysis. But mostly bad stuff happens :). This is true to some extent for anti-A and anti-B, although these are more dangerous as they fix complement in vivo better than anti-D in general. Joe Bove (my original mentor) reported a case of a patient receiving multiple units that were ABO major incompatible with no reaction. Not typical, but illustrative of the variability.

Even without a transfusion reaction, the haptoglobin drops with transfusion of red cells. Lots of non-viable cells and free hemoglobin in many red cell transfusions. If you cannot see red urine, red plasma and a drop in hematocrit/failure to rise, it's not a hemolytic reaction. Haptoglobin plays almost no role in assessing hemolytic transfusion reactions, and, as mentioned, unless you measure it on the pre-transfusion sample as well, tells you almost nothing. LDH pre and post would be more useful in most cases. Don't bother with haptoglobin in most cases.

The rate of alloimmunization in massive transfusion trauma patients who receive D positive red cells appears to be quite a bit lower than seen in healthy adults exposed to D positive red cells many years ago in the original trials. Perhaps as low as 1-2%, not 40% as seen in healthy recipients of test doses.

We only use it for antibodies that are unlikely to be clinically significant (an example would be anti-M) if they do not react at 37 and antiglobulin phase. Sometimes you find an anti-M that reacts weakly at 37, but not antiglobulin phase, and the prewarm makes it disappear at 37. In most cases, we would then ignore the antibody. Just some additional information to make clinical decisions. We wouldn't bother for antibodies usually capable of causing red cell destruction at body temperature. A strong cold with broad thermal amplitude up to 37 would not be much helped by pre-warming, so we use other techniques such as auto-adsorption to make it "go away." Thus we don't bother there with prewarming, just as we wouldn't be interested in making an anti-C or anti-Jka disappear by prewarming.

5 ml/kg plus whatever amount is needed for the dead space, although they can flush that with crystalloid. One unit minimum perhaps. That's 15 ml/kg for a 1 kg premature newborn. Would suggest a maximum of 10 ml/kg. 5 unit pool would be about 75 ml/kg for a 1 kg neonate, a terrible over dose that would increase the risk of congestive heart failure due to volume, and thrombosis due to excessive fibrinogen and factor VIII in cryo. My advice is don't do that no matter how much your neonatologists want to do that. Potentially fatal. No possible therapeutic rationale.

My condolences to her family, colleagues and friends.

I don't think the AABB comments are evidence based. Washing with 37 degree saline is extremely unlikely to cause false negatives with clinically significant antibodies, and I'm unaware of any evidence that this is so. Any such antibody would be very low affinity to be washed away by saline at any temperature, and unlikely to have in vivo/clinical significance. As argued persuasively above by Malcolm Needs, anything that doesn't react at 30 degrees or above in typical serologic testing isn't going to cause clinical problems. Patients are neither at 30 degrees nor centrifuged :). Our serologic techniques are overly sensitive, in general, for clinically insignificant agglutinins. No need for cold panels ever, with rare exception, and more for intellectual curiosity than clinical decision making. Perhaps a mini-cold screen someetimes just to confirm you are indeed detecting a weak cold agglutinin in 37 degree testing, which disappears with prewarm technique. Like Malcolm, I've never seen a patient with an hemolytic reaction due to an antibody that disappears with prewarming, in close to 50 years of clinical practice. I know there are in vitro examples of clinically significant antibodies that weaken or disappear with prewarm, but I've never seen any clinical consequences.

In emergencies, we always accept verbal orders for transfusion. These should be followed up by a request documented in our electronic medical record, but that's after the fact. If you have a paper system, then the followup order is documented that way. There is a regulatory/accreditation requirement, which I consider bureaucratic, obstructive and useless, that these emergency requests require a signed release from the ordering practitioner, if the transfusion is not fully tested for the recipient.

Agree that a urine specimen isn't indicated for an allergic reaction, so no worries.

Here's one paper that involves extended cold storage of room temperature platelets. They actually seemed more functional. Xu F, Gelderman MP, Farrell J, Vostal JG. Temperature cycling improves in vivo recovery of cold-stored human platelets in a mouse model of transfusion. Transfusion. 2013 Jun;53(6):1178-86. doi: 10.1111/j.1537-2995.2012.03896.x. Epub 2012 Sep 24. PMID: 22998069. Background: Platelet (PLT) storage at room temperature (RT) is limited to 5 days to prevent growth of bacteria, if present, to high levels. Storage in cold temperatures would reduce bacterial proliferation, but cold-exposed PLTs are rapidly cleared from circulation by the hepatic Ashwell-Morell (AM) receptor, which recognizes PLT surface carbohydrates terminated by β-galactose. We cycled storage temperature between 4 and 37°C to preserve PLT function and reduce bacterial growth. Study design and methods: Temperature-cycled (TC) human PLTs were stored at 4°C for 12 hours and then incubated at 37°C for 30 minutes before returning back to cold storage. PLTs stored at RT or at 4°C (COLD) or TC for 2, 5, and 7 days were infused into SCID mice and the in vivo recovery was determined at 5, 20, and 60 minutes after transfusion. Results: PLTs stored for 2 days in COLD had significantly lower in vivo recoveries than RT PLTs. TC PLTs had improved recoveries over COLD and comparable to RT PLTs. After 5- and 7-day storage, TC PLTs had better recoveries than RT and COLD PLTs. PLT surface β-galactose was increased significantly for both COLD and TC PLTs compared to RT. Blocking of the AM receptor by asialofetuin increased COLD but not TC PLT recovery. Conclusion: TC cold storage may be an effective method to store PLTs without loss of in vivo recovery. The increased β-galactose exposure in TC PLTs suggests that mechanisms in addition to AM receptors may mediate clearance of cold-stored PLTs.

Short periods of time (<12 to 24 hours say) at refrigerator temperatures have no known deleterious effect on platelet transfusion efficacy, so I would use them as I would use any platelet component stored at room temperature. I routinely approve this at my own institution when this happens.

The fact is that we have little to no evidence that platelet transfusion of any sort will mitigate post-pump bleeding. This is expert opinion only that has driven this practice. What we have learned in the last few years is that platelet transfusion as currently practiced (ignoring ABO for one thing) actually increases bleeding and mortality in some clinical settings. I'd rather have some oozing than be transfused with platelets empirically, both as a patient and a hematologist. With life threatening bleeding and abnormal platelet function as measured by a closure time or TEG/ROTEM, platelet transfusion makes sense and has some data driven support. Oozing, no data whatever.

Sounds like total rubbish from both a clinical and scientific viewpoint. Another instance how the administrative/legal model of reality is undermining civilization :).

I should have added that I recognize that some cardiac surgeons transfuse platelets routinely post-bypass in the hope of reducing bleeding. I suggest this is a traditional practice without the slightest shred of evidence for benefit. Purely guesswork and expert opinion, for which there is now evidence of harm. So whether you give cold or room temp platelets probably doesn't matter as (1) there is likely no benefit to either approach, and (2) there is likely equivalent harm either way. So my short answer is it doesn't matter, but that platelet transfusion to non-bleeding surgical patients likely doesn't help, and may increase the risk of thrombosis, inflammation and reduced host defenses against post-operative infection.

Why are we transfusing platelets to patients who aren’t bleeding? More likely to harm them than help them in my view.

I agree with the procedures above. But these are basic urgent communications required of any clinical service, and I wouldn't characterize them as critical values, which are emergencies. Perhaps it's just semantics :).