When is ABO-incompatible plasma safe to transfuse? Defining an empirical estimation of anti-A/anti-B in ABO incompatible plasma-containing blood products

Abstract

Background

This systematic review aims to determine the loads of anti-A and anti-B in plasma-containing blood products that can safely be transfused to adults.

Materials and methods

A comprehensive literature search was performed in PubMed, including studies dating back to 1940 until January 1^st, 2023. In total 1,869 articles were screened, describing 86 cases of hemolytic transfusion reactions (HTR) due to anti-A and/or anti-B in ABO incompatible products. Publications in English were included that reported the anti-A and/or anti-B IgM and/or IgG titers in the blood products, the volume of the transfused product and the clinical and/or laboratory signs compatible with an HTR. Publications reporting HTR in patients 18 years of age or younger, HTR after infusion of immunoglobulin drug products or after transfusion of units of red blood cell concentrates were excluded. Antibody load was defined as volume of the blood product in milliliters multiplied by antibody titer, defined by arbitrary units (AU). The studies were analyzed and evaluated to determine an amount of ABO antibodies in plasma that can be used safely for transfusion purposes.

Results

We identified HTRs with a wide-ranging antibody content, with the lowest being a platelet product with an antibody content of 12,480 AU for both IgM and IgG. However, we could establish that a calculated cutoff of 10,000 AU for IgM as well as 30,000 AU for IgG in plasma could be considered safe for transfusion.

Discussion

While titer is considered the gold standard when evaluating risk for HTR, the total volume of plasma transfused is rarely considered. Here, we recommend calculating an antibody content by using a titer and plasma volume of the unit(s) intended for transfusion as a good risk mitigation strategy for reducing HTRs.

INTRODUCTION

Blood transfusions play a crucial life-saving role in modern clinical medicine. An important part of transfusion medicine is to ensure safety and compatibility of the transfused blood products. One obstacle is the presence of antibodies against blood group antigens A and B, which can cause adverse reactions in the recipient. Major transfusion incompatibility can lead to a severe, life-threatening reaction that occurs when the recipient’s antibodies, naturally occurring anti-A and/or -B reacts against the donor’s red blood cells (RBC) causing severe acute intravascular hemolysis¹. Minor transfusion incompatibility, in comparison, is usually a mild reaction and occurs when antibodies present in the donor’s plasma are transfused. Hence, the transfusion of type O platelets into patients of blood groups other than O, are classified as a minor transfusion incompatibility. There has been increasing research during the last decade within the field of transfusion medicine on minor transfusion incompatibility. Although minor incompatibility usually causes a less severe form of transfusion reaction, there exist case reports where severe morbidity and even death have been attributed to anti-A or anti-B.

A common approach to mitigate adverse reactions due to minor incompatibility has been to establish a safe anti-A/anti-B titer cutoff for type O platelets or other type O blood components that contain donor plasma. Although the titer is used in almost all blood banks/transfusion centers as a measure of the safety of a blood product prior to a minor incompatible transfusion, there is no consensus on the manual titration methods or which cutoff titers of anti-A and anti-B to apply, and consequently, cutoff titers vary a lot around the world (Table I).

Table I.

Suggested IgM and IgG antibody cutoff titers

References	Blood product	Titer IgM	Titer IgG
Quillen et al . 2011 2	All platelet products	50	100
Cooling et al . 2008 3	Pooled platelets	128	200
Josephson et al . 2004 4	Apheresis platelets	64	256
US Armed Forces 5	Type O whole blood	250	250
Most Norwegian blood banks 6	Type O whole blood	250	500
Japan 7	ABO-incompatible HLA matched platelets	128	128
Berséus et al . 2013 8	“Universal” type O products	100	400
Germany9	Apheresis platelets	64	Not tested
Italy9	Apheresis platelets	64	256
Norway9	Apheresis platelets	Not tested	250
Sweden 9	Apheresis platelets	100	400

One drawback associated with this school of thought is that severely ill patients requiring multiple transfusions may be exposed to an increasing antibody load as the number of minor mismatch-compatible components are transfused. Therefore, a more logical way of looking at how to prevent minor incompatibility reactions should be based on defining a safe antibody load that can be transfused in total instead of relying solely on antibody titer in individual transfusion units. Additionally, as different components contain varying amounts of plasma, this would result in different antibody load to the recipient at equivalent antibody titer. Further, the blood volume of the patient should be considered: what is a safe load of anti-A/anti-B to be transfused to an average adult woman with total blood volume of 4.5 L compared with a man (5.5 L) or even pediatric patients who have a considerably less blood volume?

Objectives

We wanted to determine, by way of a systematic literature review, whether we could better estimate the risk of a hemolytic transfusion reaction (HTR) based on a cumulative score where the transfused antibody load is calculated, instead of solely focusing on the antibody titer. Here, we have performed a systematic review of papers that have reported transfusion reactions due to minor incompatibility with the overall aim of defining a safe amount of anti-A/anti-B that can be transfused without causing hemolysis in the recipient.

MATERIALS AND METHODS

Protocol and registration

This review was conducted using the PRISMA extensions for Scoping Reviews. Briefly, a twenty-point checklist was completed according to the PRISMA guidelines¹⁰.

Eligibility criteria

Studies and reports in which adult patients or volunteers (age ≥19) received a blood product containing ABO incompatible plasma (apheresis platelet units, pooled platelet units, plasma, granulocyte concentrates or whole blood units) where a transfusion reaction was reported.

Publications included in this review were case reports, case control studies, letters to the editor and data from hemovigilance reports from the UK and the USA.

Information sources

We searched PubMed for articles published from January 1^st, 1940, until January 1^st, 2023. Detailed search result queries and results can be seen in Table II. In addition, the reference lists of included papers as well as review papers that addressed the topic of hemolysis after minor mismatch transfusions were examined to find reports that had not been listed in PubMed. Moreover, all the annual reports from the UK’s Serious Hazards of Transfusion (SHOT) from 1996 until 2021 (SHOT Annual Reports and Summaries - Serious Hazards of Transfusion [shotuk.org/shot-reports]) and the annual reports from the US Food and Drug Administration (FDA) regarding transfusion-related fatalities available from 2013 until 2020¹¹ were examined for eligible cases. For cases which were both described in an abstract from a conference proceeding and in a full paper, only the latter were reported. A similar approach was used for cases which were both reported as part of a national hemovigilance program report and in a full paper. Uncertainties were resolved after discussion among all Authors.

Table II.

Search results using PubMed search engine

Search	Query	Results
#1	Search: ((((transfusion reaction[MeSH Terms]) OR (hemolysis[MeSH Terms]) AND (ABO blood group system[MeSH Terms]) OR (ABO-incompatible) OR (anti-A) OR (anti-B)))) Filters: humans, English, adult: 19+ years, from 1975/1/1 - 2023/1/1.	1,720 articles
#2	Search: (((universal donor) OR (group O blood)) OR (O negative blood) OR (group-O blood))) Filters: English, from 1940/1/1 - 1975/1/1	142 articles
#3	#1 + #2	1,862 articles

Inclusion criteria

Publications in English that reported on:

1. the anti-A and/or anti-B IgM and/or IgG titers in the blood product;

2. volume of the transfused product or a description of the blood product that was transfused, which allowed a realistic assumption of the amount of transfused plasma (see below);

3. clinical and/or laboratory signs compatible with an HTR.

Exclusion criteria

Publications reporting HTR:

1. in patients 18 years of age or younger;

2. after infusion of immune globulin drug products.

Data extraction process

From each report included, the following variables were collected and compiled into a table: author names and publication year, recipient’s age and sex, diagnosis, recipient’s ABO blood group, type of blood product received, blood group of the product, anti-A and/or anti-B titers (IgM and/or IgG), plasma volume transfused, outcome for the patient, comments regarding the specific case, clinical symptoms and/or lab results.

Volume estimation

Certain volume estimates were used due to many case reports that did not contain information of the volume of the plasma being transfused. 1) For cases with whole blood transfusions we assumed a hematocrit (Hct) of 0.40 and a total volume of the unit of 500 mL, 2) for platelet units where platelets were resuspended in plasma, we assumed a volume of 300 mL plasma, and 3) for apheresis platelets in platelet additive solution we assumed a plasma content of 1/3 of the total volume.

Quantification of antibodies

The total amount of anti-A and anti-B in each study was reported as arbitrary units AU defined as IgG and IgM titers multiplied by the volume of the transfused plasma in mL.

Titer reporting

Titers reported at room temperature in NaCl were assumed to be direct agglutination tests, and a measurement of IgM antibodies, whereas titers reported using an indirect antiglobulin test (IAT) at 37°C were considered IgG.

Hemolysis definition

For some of the older studies in which hemoglobin (Hb), hematocrit (Hct) and RBC concentration was reported, only cases where there was a reduction of ≥8% for at least two of these variables within 48 hours were categorized as HTR^12,13. We reasoned that an 8% decrease of these variables is equivalent to the loss of one unit of RBCs during blood donation and thus would represent the equivalent loss due to hemolysis.

RESULTS

Case selection

The titles of 1,862 articles were screened for eligibility. Abstracts were further examined for potentially relevant papers and finally the full manuscripts were scrutinized for fulfilling the inclusion and exclusion criteria. Likewise, we examined reference list of included papers as well as review papers that addressed the topic of hemolysis after minor mismatch transfusions. Eleven reports fulfilling the inclusion criteria were retrieved from scrutiny of reference lists^14–24 (Figure 1).

Figure 1.

An overview of the included reports

*: Seven papers were found by scrutinizing reference lists of papers that are included in the review. Four papers were not found in PubMed¹⁵–¹⁸. Three papers were found in PubMed but did not come up on any of the searches¹⁴,¹⁹,²⁰.

Synthesis of results

We identified a total of 43 papers published between the years of 1940 and 2023 that contained 90 cases where transfusion of blood products containing anti-A and/or anti-B to patients of non-group O patients had resulted in HTR. The results of the papers included are compiled in Table III.

Table III.

Minor incompatible transfusion cases that resulted in HTR

First author, yearRef	Recipient’s age, sex and outcome	Diagnosis	Recipient’s ABO group/blood product’s ABO group	Type of blood product	Plasma volume [mL]	Anti-A or anti-B IgM titer	Anti-A or anti-B IgM content [volume in mL × titer]	Anti-A or anti-B IgG titer	Anti-A or anti-B IgG content [volume in mL × titer]	Comments	Clinical signs and laboratory values
Aubert, 1942 14	20 M -	Healthy volunteer	A/O	Serum	500	1,024	512,000	-	-		Slight headache, mild “muscular” back pain, bilirubin 0.8 → 4.0 mg/dL
	24 F -	Healthy volunteer	A/O	Serum	500	2,048	1,024,000	-	-		Temp 37.5°C, subjective symptoms, bilirubin 0.5 → 4.0 mg/dL
	54 F -	Hematemesis	A/O	Serum	150	8,192	1,228,800	-	-		Pins-and-needles feeling in both legs, hemoglobinemia, severe back pain, hemoglobinuria, bilirubin 0.4 → 3.8 mg/dL
	23 F -	Post partum hemorrhage	A/O	Serum	350	2,048	716,800	-	-		Rigor, bilirubin 0.2 → 2.2 mg/dL
	20 F -	Pyelitis, pregnancy	A/O	Serum	30	8,192	245,760	-	-		Neck pain, breathing problems, bilirubin 0.5 → 2.7 mg/dL, BP 110/70 → 140/70 mmHg
	43 F -	Menorrhagia	A/O	Serum	130	8,192	1,064,960	-	-		Back pain, jerky spasms, nausea, temp 38.3°C, bilirubin 0.0 → 1.7 mg/dL
	38 F -	Uterus prolapse	A/O	Serum	450	16,384	7,372,800	-	-		Temp 38.3°C, nausea, restlessness, bilirubin 0.2 → 3.2 mg/dL
	46 F -	Carcinomatosis	A/O	Serum	400	512	204,800	-	-		Knee pain, nausea, hemoglobinemia, bilirubin 0.2 → 2.4 mg/dL
	32 M -	Left subphrenic abcess	A/O	Plasma	250	1,024	256,000	-	-	2 weeks later given 350 mL serum (titer anti-A 32) without problems	Knee pain, temp 38.3°C, Hb decrease 24%, hematocrit decrease 12%
	57 M -	Carcinoma colon	A/O	Serum	225	2,048	460,800	-	-	Was later given 450 mL serum (titer anti-A 128) without problems	Pain (chest, loins, groins, thighs, and arms), rigor, hemoglobinemia, bilirubin 0.1 → 1.33 mg/dL, Hb decrease 5%, hematocrit decrease 3% (although these changes were in 3 days)
	51 F -	Gastric ulcer	A/O	Plasma	450	512	230,400	-	-		Back pain, cold and shivering, nausea, temp 38.8°C, bilirubin 0.1 → 0.53 mg/dL, RBC decreased 0.75 million
Weintraub, 1943 15	23 M -	Hypoplastic anemia	A+/O	Whole blood	300	320	96,000	-	-	Estimated volume (0.5 L × 0.6)	Severe chills, temp 40.6°C, hemoglobinuria, bilirubin 0.6 mg/dL, hematocrit 41 → 35%
Wiener, 1943 16	- F Died	Preeclampsia	A-/O	Whole blood	240	512	122,880	-	-	Estimated volume (0.4 L × 0.6)	Severe back pain, chill, temp 40.6°C, jaundice, anuria
Malkiel, 1945 17	19 F -	-	A/O	Whole blood	270	1,024	276,480	-	-	Estimated volume (0.45 L × 0.6)	Feeling of “oppression”, severe chill, temp 38.9°C, jaundice, oliguria
Alberton, 1945 18	77 F Died	Gastric cancer	A/B	Plasma	100	900§	90,000	-	-	The patient was transfused with a mixture of plasma and saline	Severe back pain, chill, temp 38.3°C, hemoglobinemia, hemoglobinuria
Morgan, 1946 25	38 M Died	Abdominal wound	A1+/O	Whole blood	340	258	87,720	-	-	Estimated volume (0.57 L × 0.6)	Hb 85%, hemoglobinuria, oliguria, kidney failure
			A1+/O	Whole blood	340	4,096	1,392,640	-	-
			A1+/O	Whole blood	340	512	174,080	-	-
Tisdall, 1946 12	- M -	Healthy “volunteer”£	A-/O	Plasma	250	500	125,000	-	-		Hb ↓ 25%, RBC ↓ 12%, urticaria
	- M -	Healthy “volunteer”£	B+/O	Plasma	250	500	125,000	-	-		Hb ↓ 16%, hematocrit ↓ 11%, RBC ↓ 11%
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	500	125,000	-	-		Hematocrit ↓10%, RBC ↓ 21%, urticaria
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	600	150,000	-	-		No significant reduction in Hb, hematocrit or RBC, low back pain, hemoglobinuria
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	750	187,500	-	-		Hb ↓ 20%, RBC ↓ 19%
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	750	187,500	-	-		Hb ↓ 16%, hematocrit ↓ 16%, RBC ↓ 12%
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	750	187,500	-	-		Hb ↓ 10%, hematocrit ↓ 15%, urticaria
	- M -	Healthy “volunteer”£	A+/O	Plasma	240	800	192,000	-	-		Hb ↓ 11%, hematocrit ↓ 11%
	- M -	Healthy “volunteer”£	AB+/O	Plasma	250	1,000	250,000	-	-		Hematocrit ↓ 11%, hemoglobinuria
	- M -	Healthy “volunteer”£	A-/O	Plasma	250	1,000	250,000	-	-		Hb ↓ 13%, hematocrit ↓9%
	- M -	Healthy “volunteer”£	A-/O	Plasma	250	1,000	250,000	-	-		Hb ↓ 9%, RBC ↓ 9%
	- M -	Healthy “volunteer”£	A-/O	Plasma	250	1,000	250,000	-	-		Hematocrit ↓ 10%, RBC ↓ 10%
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	1,000	250,000	-	-		Hb ↓ 8%, hematocrit ↓ 11%
	- M -	Healthy “volunteer”£	B+/O	Plasma	250	1,000	250,000	-	-		Hb ↓ 11%, RBC ↓ 11%
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	1,250	312,500	-	-		Hb ↓ 21%, hematocrit ↓ 13%
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	1,500	375,000	-	-		Hematocrit ↓ 13%, hemoglobinuria
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	1,500	375,000	-	-		Hematocrit ↓ 11%, RBC ↓ 12%
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	2,000	500,000	-	-		Hematocrit ↓ 8%, RBC ↓ 9%, bilirubin 0.7 → 3.1 mg/dL, nausea, vomiting, sweating, hemoglobinuria
	- M -	Healthy “volunteer”£	A+/O	Plasma	200	2,500	500,000	-	-		No significant reduction in Hb, hematocrit, or RBC, bilirubin 0.35 → 4.0 mg/dL, hemoglobinuria, back pain, chills, temp 38.4°C
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	2,500	625,000	-	-		Hb ↓ 21%, hematocrit ↓ 18%, back pain, chills
	- M -	Healthy “volunteer”£	A+/O	Plasma	120	3,000	360,000	-	-		Hematocrit ↓ 9%, RBC ↓ 9%, back pain, nausea, headache, temp 38.1°C
	- M -	Healthy “volunteer”£	B+/O	Plasma	250	3,000	750,000	-	-		Hb ↓ 13%, hematocrit ↓ 11%, bilirubin 0.8 → 1.95 mg/dL, chills, temp 38.1°C
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	3,000	750,000	-	-		Hb ↓ 11%, RBC ↓ 9%, bilirubin 0.8 → 1.95 mg/dL, hemoglobinuria, chills, temp 38.8°C
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	3,000	750,000	-	-		No significant reduction in Hb or hematocrit, bilirubin 0.45 → 2.05 mg/dL, hemoglobinuria, chills
	- M -	Healthy “volunteer”£	A+/O	Plasma	250	4,000	1,000,000	-	-		No significant reduction in Hb, hematocrit or RBC, bilirubin 0.35 → 1.65 mg/dL, hemoglobinuria, headache
Tisdall, 1946 13	- M -	Healthy “volunteer”£	A1/B	Plasma	25	2,400	60,000	-	-		No significant reduction in Hb, hematocrit or RBC, bilirubin 1.10 → 2.10 mg/dL, temp 37.4°C, hemoglobinuria, chills, back pain
	- M -	Healthy “volunteer”£	A1/B	Plasma	50	2,400	120,000	-	-		No significant reduction in Hb, hematocrit or RBC, bilirubin 0.60 → 2.38 mg/dL, hemoglobinuria, back pain
	- M -	Healthy “volunteer”£	A2/B	Plasma	50	1,500	75,000	-	-		No significant reduction in Hb, hematocrit or RBC, bilirubin 0.87 → 1.40 mg/dL, hemoglobinuria, chills
Ebert, 1946 26	- Survived	Gunshot wound	A+/O	Whole blood	210	2,048	430,080	-	-	Estimated volume (0.35 L × 0.6) Presumably adult patient	Back pain, nausea, vomiting, chilliness, hemoglobinuria, hemoglobinemia, bilirubin 1.6 → 3.7 mg/dL
Ervin, 1950 27	- Survived	Healthy volunteer	A1/O	Plasma	250	1,280	320,000	1,280	320,000	Presumably adult	Hemoglobinemia 240 mg/dL, hemoglobinuria, chilliness, moderate fever, hematocrit 42 → 24%, bilirubin 0.7 mg/dL, DAT pos
	- -	Gastric carcinoma	A1/O	Whole blood	300	320	96,000	40	12,000	Estimated volume (0.5 L × 0.6) Presumably adult patient	Jaundice, hemoglobinuria, bilirubin 16.8 mg/dL
	- -	Supracondylar amputation	A1/O	Whole blood	300	10,240	3,072,000	2,560	768,000	Estimated volume (0.5 L × 0.6) The patient received 30 mL of A and B substance	Jaundice, hemoglobinuria, bilirubin 2.6 mg/dL
Grove-Rasmussen, 1953 19	67 M -	Meningioma	A+/O+	Whole blood	300	64	19,200	1,024	307,200	Estimated volume (0.5 L × 0.6)	Hematuria, Hb 13.6 → 8.6g/dL, bilirubin 7.5 mg/dL, DAT pos
Buchanan, 1953 28	67 F Died	Carcinoma of vulva	A+/O+	Whole blood	163	16	2,608	-	-	Two transfusions of whole blood; 500 mL each → 163 mL of plasma in each unit	Hematuria, DAT pos, oliguria
					163	256	41,728
Elliott, 1954 29	78 F Died	Biliary colic	A1+/O+	Whole blood	110	32,768	3,604,480	-	-	250 mL whole blood → 110 mL plasma	Severe rigors, pallor, sweating, shock, disorientation, temp 38.3°C,jaundice, bilirubin 3.6 mg/dL, 4th day: BP 220/110 mmHg, bilirubin 11.2 mg/dL, oliguria → coma → death
Stevens, 1954 30	39 M Survived	Healthy volunteer	A/O+	RBC unit	210	128	26,880	5,120	1,075,200	Estimated volume (0.35 L × 0.6)	Flushing, tingling of the hands, chills, temp 38.3°C, lumbar back pain, BP 130/80 → 110/78 mmHg, hemoglobinemia, hemoglobinuria, oliguria, DAT pos
Kendrick, 1964 21	- M -	Abdominal wound	A/O	Whole blood	45	8,000	360,000	-	-	Estimated volume (0.075 L × 0.6)	Severe chills, temp 40.6°C, bilirubinemia, jaundice
Ballas, 1968 22 Barnes, 1973 31	- Survived	Multiple fractures, amputation	A/O-	Whole blood	270	256	69,120	32,768	8,847,360	Estimated volume (0.45 L × 0.6)	Oliguria, hematocrit 32 → 20%, bilirubin 1.9 mg/dL
Chipping, 1980 32	45 F Survived	Preleukemia	A+/O+	Granulocyte concentrate	200	-	-	512	102,400	Assuming the 2nd leucocyte infusion was also 200 mL, the total volume of plasma was estimated to 200 mL (0.40 L × 0.5)	Rigor, nausea, hemoglobinuria, Hb 9.0 → 5.8 g/dL, DAT pos
Inwood, 1978 33	31 F Survived	CML	A-/O	RBC unit	95	-	-	8,192	778,240		Rigor, vomiting, Hb 6.3 → 4.4 g/dL, bilirubin 0.5 → 1.8 mg/dL, haptoglobin 150 → 25 mg/dL, DAT 3+, hemoglobinuria, hemoglobinemia
Thomas, 1981 34	44 M Survived	AML	A-/O	Platelet apheresis	300	16	4,800	16	4800	Estimated volume. The units were transfused over a period of 4 days	No clinical signs of hemolysis, Hb 8.6 → 7.6 g/dL, bilirubin 3.9 → 10.3 mg/dL
			A-/O	Platelet apheresis	300	32	9,600	256	76,800
			A-/O	Platelet apheresis	300	32	9,600	256	76,800
McLeod, 1981 35 McLeod, 1982 36	45 M Survived	Erythroleukemia	A1/O	4 single plt. units	200	-	-	5,120	1,024,000	Average titer of the 4 units (10,240×2/4)	Hb 14 → 8 g/dL, bilirubin 9.9 → 48 mg/dL, hemoglobinuria, DAT pos, creatinine 1.6 → 6.0 mg/dL
Siber, 1982 37	20 M Survived	AML	A+/O+	Platelet apheresis	199	-	-	8,192	1,630,208	Platelet donor received vaccine 4 weeks earlier	Hematocrit 22 → 16%, haptoglobin 420 → 0 mg/dL, DAT pos, temp 38.3°C, hemoglobinuria, hemoglobinemia, chill
Pierce, 1985 38	58 F Survived	Acute bypass operation	B+/O	1 single plt. unit	50	512	25,600	16,384	819,200	Plasma volume was 50–70 mL	Hb 14.3 → 8.2 g/dL, bilirubin 3.2 → 7.0 mg/dL, haptoglobin 95 → 35 mg/dL, DAT 2+, LDH 770 U/L
Ferguson, 1988 39	66 M Survived	AML	A-/O	1 single plt. unit	50	256	12,800	>4,000	>200,000		Hb decrease 2.6 g/dL, DAT 3+, dyspnea, back pain, flushing, chills, hemoglobinuria
Reis, 1989 40	56 M Survived	Aplastic anemia	B+/O+	Platelet apheresis	300	-	-	4096	1,228,800	Estimated volume. Titer of donor’s plasma	Hb 11.3 → 5.2 g/dL, bilirubin 1.5 mg/dL, haptoglobin 20 mg/dL, DAT pos, fever, malaise, weakness, light-headedness, vomiting, dyspnea, and dark urine
Murphy, 1990 41	30 F Survived	AML	A1/O	Platelet apheresis	448	256	114,688	1,024	458,752	Titer of donor’s plasma	Hb 11.4 → 6.0 g/dL, DAT pos, pale, icteric
Mair, 1998 42	28 M -	Neuroblastoma	A/O	Platelet apheresis	225	128	28,800	-	-		Hb 8.4 → 5.8 g/dL, bilirubin 1.1 → 3.3 mg/dL, LDH 170 → 372 U/L, dyspnea, chills, lower back pain, BP 186/106 → 136/79 mmHg, heart rate 76 → 88, temp 36.7 → 37.3°C, DAT pos
SHOT, 1998- 1999 43	36 F Survived	Post partum hemorrhage	A+/O	Platelet apheresis	300	-	-	20,000	6,000,000	Estimated volume	DAT pos, evidence of intravascular hemolysis
Larsson, 2000 44	44 F Survived	AML	A/O	Platelet apheresis	371	16,384	6,078,464	-	-	Drop in Hb despite transfusion of 2 units of RBCs of blood group A	Hb 7.7 → 5.4 g/dL (despite transfusion of 2U RBC), bilirubin 0.4 → 13 mg/dL, LDH 431 → 2.485 μ/L, DAT 3+, back pain, tingling of the fingers, shortness of breath, flushed behind ears
Valbonesi, 2000 45	51 F Survived	Metastatic breast cancer	A+/O+	Platelet apheresis	35	>8,000	>280,000	-	-	Titer of donor’s plasma Hb and bilirubin values were obtained from Figure 1A in ref. No. 45.	Hb 8.5 → 5.2 g/dL, bilirubin 0.5 → 6 mg/dL, LDH 488 → 2.969 IU, DAT pos, hypotension (BP 80/40 mmHg), shock
Sauer-Heilborn, 2002 23	35 - Survived	SCT	B+/O-	Platelet apheresis	526	4,096	2,154,496	2048	1,077,248		Hematocrit 32 → 16%, bilirubin 6 → 13.8 mg/dL, fever, hemoglobinuria, acute renal failure, hypotension, DAT 2+
Gresens, 2003 46	29 M Survived	Gunshot wound	A/O	Platelet apheresis	200	1,024	204,800	1024	204,800	Estimated volume: ”After most of the 260 mL unit had been transfused...”	Transient hypotension, temp 36.6 → 37.9°C, DAT 3+
Barjas-Castro, 2003 47	38 M -	HTLV-1-associated non-Hodgkin’s lymphoma	A1-/O-	RBC unit	55	1,024	56,320	1024	56,320		Temp 38.3°C, dyspnoea, chills, back pain, haptoglobin 5.8 mg/dL, Hb 6.9 → 6.8g/dL (despite transfusion of 1U RBC), LDH 765 → 1,636 U/L, DAT 2+
SHOT, 2003 48	31 M -	ALL	A1/O	Platelet apheresis	300	1,024	307,200	8192	2,457,600	Estimated volume	Hemolysis, DAT+, hyperbilirubinemia, spherocytosis, renal failure, Hb decrease
Fauzie, 2004 24	Adult	APL	A/O	Platelet apheresis	598	256	153,088	512	306,176		Hb decrease 1–2 g/dL, DAT+, cyanosis, dyspnea, hemoglobinuria
	Adult -	ALL	A/O	Platelet apheresis	390	32	12,480	32	12,480		Hb decrease 1–2 g/dL, DAT+, severe back pain
Josephson, 2004 4	Adult -	Acute leukemia	A/O	Platelet apheresis	50	256	12,800	8,196	409,800		Rigors, severe back pain, hemoglobinuria, hemoglobinemia, Hb decrease 3 g/dL, DAT+
	Adult -	Acute leukemia	A/O	Platelet apheresis	50	-	-	1,024	51,200		Rigors, severe back pain, hemoglobinuria, hemoglobinemia, Hb decrease 3 g/dL, DAT+
Zubair, 2004 49	67 M -	Mantle cell lymphoma	A/O	Platelet apheresis	150	512	76,800	-	-		Nausea, light-headiness, chills, back pain, bradycardia and hypotension, Hb 10.6 → 7.4 g/dL and total serum bilirubin increased by 9 fold. DAT 1+
Sadani, 2006 50	65 F Survived	AML	A+/O	Platelet apheresis	300	128	38,400	1,280	384,000	Estimated volume	Hemoglobinuria, jaundice, Hb 7.7 → 3.9 g/dL, hematocrit 24 → 12%, billirubin 21 → 68 μmol/L, LDH 124 → 2.157 U/L, urea 4.7 → 22.3 μmol/L, creatinine 43 → 131 μmol/L, DAT 3+
Cooling, 2007 3	36 -	Hodkin’s lymphoma	A/O	Platelet apheresis	214	-	-	2,048	438,272	Data from Table I in ref. No. 3	Hb decrease 2.5 g/dL (22%)
	45 - -	AML	A/O	Platelet apheresis	241	-	-	4,096	987,136	Data from Table I in ref. No. 3	Hb decrease 1.6 g/dL (15%)
Daniel- Johnson, 2009 51	40 M Survived	AML	B+/A+	Platelet apheresis	100	16,384	1,638,400	16,384	1,638,400	Titer of donor’s plasma	Severe back flank, chest pain, BP 112/72 → 150/92 mmHg, Hb 9.7 → 5.5 g/dL, bilirubin 1 → 8 mg/dL, DAT+ Temp 37.4 → 38.5°C, BP 110/70 → 160/80 mmHg, heart rate 92 → 140,
Fontaine, 2012 52	63 F Survived	AML → SCT	O → A¥	Platelet apheresis	231	512	118,272	2,048	473,088		respiratory rate 18 → 30, Hct. 20.6 → 14.6%, DAT pos., bilirubin 0.6 → 5.6 mg/dL
Karafin, 2012 53	21 M -	MDS	B+/A+	Platelet apheresis	194	-	-	16	3,104	Received two units of apheresis platelets	Temp 36.6 → 38.6°C, Hb decrease 2.2 g/dL, DAT pos
					193	-	-	2	386
	32 F -	ALL	B+/O+	Platelet apheresis	515	-	-	64	32,960		Temp 38.5 → 37.9°C, Hb decrease 1.9 g/dL, chills, DAT neg
	41 F -	AML	AB+/O+	Platelet apheresis	313	-	-	2,048	641,024	Received two units of apheresis platelets	Temp 37.4 → 38.4°C, Hb decrease 1.7 g/dL, chills, DAT pos
					244	-	-	128	31,232
Shachner, 2018 20	53 F Died	Adenocarcinoma	B+/A+	Platelet apheresis	135	512	69,120	512	69,120		Severe back pain, Hb decrease 1.5 g/dL, DAT 1+
FDA, 2018 54	- Died	-	AB+/O	Platelet apheresis	300	-	-	32,000#	9,600,000#	Estimated volume
Balbuena-Merle 2019 55	60 F Died	AML	B+/O	Platelet apheresis	122	128	15,616	512	62,464	367 mL platelet suspension in PAS → 122 mL plasma	Lethargic, disoriented, temp 38.4°C, tachycardia (heart rate 124 beats/min), BP decrease to 79/48 mmHg, oliguria, Hb 8.2 → 4.8 g/dL, haptoglobin 324 → 109 mg/dL, DAT 3+, LDH increased to 10,000 U/L, bilirubin increased to 18.8 g/dL, creatinine increased to 1.54 mg/dL
Moinuddin, 2019 56	61 F Survived	AML	A/O	Platelet apheresis	280	512	143,360	-	-		Chills, pallor, body aches, hemoglobinuria, hemoglobinemia, heart rate 74 →81, Hb 8.1 → 4.5 g/dL, bilirubin 0.8 → 2.4 mg/dL, LDH 369 → 1,125 U/L, DAT 4+
* Guarente, 2019 57	73 M -	MDS	A+/O+	Pool of 5 single plt. unit	300	-	-	64†	19,200	Titers, antibody content and estimated volume of the platelet pool†	Rigors, decreasing O2 saturation, heart rate 64 → 82 beats/min, temp 36.8 → 39.3°C, Hb 8.3 → 6.4 g/dL, bilirubinemia 0.8 mg/dL, LDH 836 IU/L, haptoglobin <10 mg/dL, hemoglobinuria, DAT pos
					60	>200‡	>12,000	2,048‡	122,880	Titers, antibody content and estimated volume of unit from 1 of the 5 donors‡
Gammon, 2021 58	65 F -	Lymphoma	A/O	Pool of 5 single plt. unit	300	256€	76,800	2,048€	614,400	Estimated volume, average titer of the whole pooled bag	Temp 39.4°C, rigors, dyspnea, cough, chest pressure, nausea, diarrhea, hypotension and tachycardia, Hb 8.9 → 7.5 g/dL, LDH 996 U/L; haptoglobin <10 mg/dL, prothrombin time 18.3 sec, INR 1.5, APTT 47.3 sec, lactate 7.0 mmol/L, dark urine

^£The “volunteers” were recruited from the Colorado State Penitentiary, Canon City, Colorado, USA.

^§The titer reported in the paper was 2,048, but as the titer of plasma-saline mixture was found to be 900 at another laboratory we selected the lowest of these two titers.

^$In this paper, the unit for the bilirubin values was reported in “g/dL”, which probably is a typo. For this reason, we have chosen to report the bilirubin values from this paper in “mg/dL”.

^¥The blood group changed from O to A after a stem cell transplantation.

^#It is not specified whether this was IgG or IgM titer.

^*In this paper, both titers of blood product^† and the donor^‡ were measured. However, since it is difficult to find an association between these two sets of titers, both sets of data are reported.

^€Titers from a “simulated pool”.

Although most cases of HTR were found in patients who received blood products with very high antibody content (Table III) there were also cases where transfusion of a large dose of anti-A and/or anti-B did not result in HTR (Table IV). The lowest antibody load implicated in an HTR during a single transfusion was a unit of apheresis platelets with anti-A IgM and IgG titers of 32 and volume of 390 mL, which gives an antibody load of 12,480 AU for both IgM and IgG²⁴. The lowest absolute titer resulting in a transfusion reaction was an apheresis platelet unit with a plasma volume of 193 mL and an anti-A IgG titer of 2 giving an antibody content of 386 AU; the same patient received also another apheresis unit with an anti-A IgG titer of 16 and a plasma volume of 194 mL giving on antibody content of 3,104 AU⁵³. This case sticks out as an exception due to the titers being so low, however the IgM titers of these units were not reported. At the other end of the scale, one case reported an IgM anti-A titer of 32,768 AU in a 250 mL unit of whole blood, with an estimated volume of 110 mL of plasma transfused, equivalent to 3.6 million AU of IgM; the IgG titer was not reported. The patient unfortunately died by an HTR followed the transfusion of this massive amount of antibodies²⁹.

Table IV.

Minor incompatible transfusion cases that did not result in HTR

First author, YearRef	Recipient’s age and sex	Diagnosis	Recipient’s ABO group/ blood product’s ABO group	Type of blood product	Plasma volume [mL]	Anti-A or anti-B IgM titer	Anti-A or anti-B IgM content [volume in mL × titer]	Anti-A or anti-B IgG titer	Anti-A or anti-B IgG content [volume in mL × titer]	Comments	Clinical signs and laboratory values
Aubert, 1942 14	23 F	Volunteer	A/O	Serum	500	1,024	512,000	-	-		No significant change in lab values, no clinical symptoms
	32 M	Left subphrenic abscess	A/O	Plasma	350	32	11,200	-	-		No significant change in lab values, no clinical symptoms
	57 M	Carcinoma colon	A/O	Serum	450	128	57,600	-	-		No significant change in lab values, no clinical symptoms
	48 F	Menorrhagia	A/O	Serum	450	32	14,400	-	-		No significant change in lab values, no clinical symptoms
	24 F	Cesarean section	A/O	Serum	300	32	9,600	-	-		No significant change in lab values, no clinical symptoms
	54 M	Carcinoma of prostate	A/O	Serum	450	128	57,600	-	-		No significant change in lab values, no clinical symptoms
Ebert, 1946 26										4 healthy volunteers were transfused
Tisdall, 1946 12	- M	Healthy volunteer soldiers	A or B/O	Plasma	250–500	500–4,000	125,000	-	-		In 3 no clinical signs of hemolysis, temp increase to 37.4°C in 1 individual
	- M	Healthy “volunteer”£	A-/O	Plasma	250	400	100,000	-	-		No significant change in lab values, urticaria
	- M	Healthy “volunteer”£	A+/O	Plasma	250	500	125,000	-	-		No significant reduction in Hb, hematocrit and RBC, no clinical symptoms
	- M	Healthy “volunteer”£	B+/O	Plasma	250	500	125,000	-	-		No significant reduction in hematocrit and RBC
	- M	Healthy “volunteer”£	A+/O	Plasma	250	500	125,000	-	-		No significant reduction in Hb, hematocrit and RBC, no clinical symptoms
	- M	Healthy “volunteer”£	B+/O	Plasma	250	750	187,500	-	-		No significant reduction in Hb, hematocrit and RBC, chills, temp 38.1°C
	- M	Healthy “volunteer”£	A+/O	Plasma	250	750	187,500	-	-		No significant reduction in Hb, hematocrit and RBC, no clinical symptoms
	- M	Healthy “volunteer”£	A+/O	Plasma	250	750	187,500	-	-		No significant reduction in Hb, hematocrit and RBC, no clinical symptoms
	- M	Healthy “volunteer”£	A+/O	Plasma	250	1,000	250,000	-	-		No significant reduction in Hb and hematocrit, no clinical symptoms
	- M	Healthy “volunteer”£	A+/O	Plasma	275	1,000	275,000	-	-		No significant reduction in Hb or hematocrit
	- M	Healthy “volunteer”£	A+/O	Plasma	250	2,000	500,000	-	-		No significant reduction in Hb, hematocrit or RBC
	- M	Healthy “volunteer”£	A+/O	Plasma	250	2,000	500,000	-	-		No significant reduction in Hb, hematocrit and RBC, chills, temp 37.9°C
	- M	Healthy “volunteer”£	B+/O	Plasma	250	2,000	500,000	-	-		No significant reduction in Hb and hematocrit, nausea, headache
	- M	Healthy “volunteer”£	A+/O	Plasma	250	3,000	750,000	-	-		No significant reduction in Hb or RBC
	- M	Healthy “volunteer”£	A+/O	Plasma	250	3,000	750,000	-	-		No significant reduction in Hb or RBC, headache
Tisdall, 1946 13	- M	Healthy “volunteer”£	A2/B	Plasma	25	2,000	50,000	-	-		No significant reduction in Hb or RBC, bilirubin 0.80 → 1.20 mg/dL, temp 37.8°C, chills
Ervin, 1950 27	-	Healthy volunteer	B/O	Plasma	450	4	1,800	-	-	4 patients received incompatible units. Estimated volume. The titers were from later case where the platelet transfusion caused hemolysis	No significant change in lab values, no clinical symptoms
Fontaine, 2012 52	-	-	-	Platelet apheresis	300	512	153,600	2,048	614,400
Karafin, 2012 53	40–66 13M/6F	17 oncology patients	-	Platelet apheresis	220–365	-	-	512– 1,024	112,640	17 patients received 19 incompatible units	Hb decrease 0.5–1.0 g/dL
	52 M	Granulocytic sarcoma	B-/O+	Platelet apheresis	385	-	-	1,024	394,240		Temp 36.6 → 38.2°C, Hb decrease 0.3 g/dL, chills, DAT neg
Balbuena-Merle, 2019 55	-	-	B/O	Platelet apheresis	122	128	15,616	512	62,464

^£The “volunteers” were recruited from the Colorado State Penitentiary, Canon City, Colorado, USA.

Even though the character of the data did not allow for conduction of a meta-analysis, we created two scatter plots (Figure 2 and 3) showing all HTRs included in the study with all known parameters such as IgM or IgG titers, the known or estimated plasma volume, as well as commonly used antibody titer cutoffs and antibody content lines. These two figures depict the association between antibody titers, antibody contents and HTR. For Figure 2A, showing cases with IgM, we chose the commonly used IgM cutoffs of 50, 64, 128 and 250 (Table I), while for Figure 2B, showing cases with IgG, we chose the commonly used cutoffs of 100, 128, 256, and 500 (Table I). Had an IgM cutoff of 10,000 AU been applied, all HTR could have been avoided except for 3 cases. By applying a very low IgM cutoff titer of 50, four cases of HTRs would have been missed (Figure 2A). Similarly for IgG (Figure 2B), using 30,000 AU as a cutoff would have resulted in six missed cases of HTR, whereas an IgG cutoff titer of 100 would have resulted in seven missed cases of HTR (Figure 2B).

Figure 2.

(A) Scatter plot showing all cases collected with IgM titer and plasma volume, three commonly used IgM cutoffs of 64, 128, and 250, as well as two lines depicting an IgM antibody content of 10,000 AU and 20,000 AU. (B) Scatter plot showing all cases collected with IgG titer and plasma volume, three commonly used IgG cutoffs of 128, 256, and 500, as well as two lines depicting an IgG antibody content of 30,000 AU and 50,000 AU.

Figure 3.

(A) Scatter plot showing the cases where patients (color coded) received multiple units of blood products where the IgM titer is known of the individual units with commonly used titer cutoffs and antibody content lines. (B) Scatter plot showing cases from Figure 3A while also showing the antibody content of each transfused unit, the total antibody content for each case, as well as antibody content cutoffs of 10,000 AU and 20,000 AU.

There were a few cases where the patients received multiple units of blood products containing minor ABO-incompatible plasma (Figure 4)^{25,28,34,53,57}. Figure 4A shows the cases with patients receiving multiple units where only IgM or IgG titers respectively were known, with commonly used titer cutoffs as well as the suggested AU content lines. In Figure 4B, we have plotted the same cases from 4A, while showing the AU content of corresponding units in the same colors, the sum of all AU contents in a different color and the AU content cutoff lines.

Figure 4.

(A) Scatter plot showing cases in which patients (color coded) received multiple units of blood products where the IgG titer is known of the individual units with commonly used titer cutoffs and antibody content. (B) Scatter plot showing cases from Figure 4A while also showing the antibody content of each transfused unit, the total antibody content for each case, as well as antibody content cutoffs of 30,000 AU and 50,000 AU

DISCUSSION

The issue regarding the clinical consequences of minor mismatched transfusions has been a frequently debated topic for decades within transfusion medicine^4,42,59,60. Many early studies, conducted during the 1940s until the 1960s, focused on the safety of minor incompatible type O whole blood transfusions, and the enigma of exactly defining safe titer for the so-called “universal donors” ^{15–17,19,22,25,27,29,31}. However, even with the widespread use of a safe titer cutoff, there is still no current consensus on what exactly that cutoff should be⁶¹. There currently exists a wide range of policies at blood banks across the world to prevent HTRs due to transfusion of products containing incompatible plasma, mostly due to a shortage of blood products of the same blood group. A summary of the different titers suggested to be safe are presented in Table I.

However simple titers do not tell the whole story. In current blood bank practice, it is not possible to add up titers of multiple transfused units to make a risk assessment of a potential HTR, so the titers of each unit must be considered individually. Therefore, we reasoned that if the sum of the antibody content of all transfused units was calculated, it would give a better estimate of risk of an HTR (shown in Figure 4). We show in this systematic review that the majority of cases can be avoided if we use our suggested cutoffs of 10,000 AU for IgM and 30,000 AU for IgG as the sum of all transfused units. It must be noted that due to so few cases with multiple units being transfused that resulted in HTRs, it is hard to make an accurate extrapolation, however the collected data does show that the approach of using a sum of antibody content transfused is plausible.

We did not observe a linear relationship between antibody load and the occurrence of HTR (Figures 2 and 3). It has also been observed that high amounts of antibodies do not necessarily result in HTR (Table IV). Certain patient-specific factors might contribute to this, such as the existence of ABO antigens on other cells. It is known that endothelial cells express ABO antigens on their surface that can bind anti-A and anti-B^60,62.

In the context of HTR, secretor status can be a contributing factor to why some individuals may experience HTRs at a certain antibody concentration while others do not. Secretors may have higher amounts of soluble ABO antigens in their blood, potentially leading to a greater ability to neutralize transfused incompatible antibodies and thereby decreasing the risk of HTRs³¹.

The activation of complement pathways can lead to hemolysis. Variability in the activity and regulation of complement pathways among patients can contribute to differences in HTR susceptibility. It is also important to take into consideration the density of A and B antigens on RBCs, as this can vary among individuals^63,64. When antibodies encounter RBCs with a higher antigen density, more antibodies will bind to the RBC which, in turn, can lead to stronger complement activation due to a greater number of antigen-antibody complexes. The RBC membrane is equipped with complement regulatory proteins, such as CD55 (decay-accelerating factor) and CD59 (protectin), which serve to inhibit complement activation^65,66. Variation in the density of complement inhibitors on the RBC surface, such as CD55 and CD59, and in the concentration of plasma complement inhibitors, such as Factor I and Factor H, may lead to decreased complement activation.

Antibody characteristics can also be important as the specificity of anti-A and anti-B can vary. Some antibodies may have higher affinity for target antigens, making them more likely to cause HTRs, while others may be less effective in binding and causing reactions. IgM antibodies are known to be potent activators of complement while the different IgG subclasses also vary in their ability to activate the complement system. IgG1 and IgG3, for example, are more likely to activate complement, whereas IgG2 and IgG4 are less potent⁶⁷. It has been suggested previously that IVIg-associated hemolysis involves a two-hit mechanism⁶⁸. The passive transfer of ABO hemagglutinins constitutes the first hit, while the inflammatory state of the recipient constitutes the second hit, resulting in accelerated removal of the sensitized RBCs. This two-hit mechanism is very similar to the proposed model for TRALI which is mediated by HLA/HNA/bioactive lipids⁶⁹. As in the TRALI model, where a threshold effect has been proposed⁷⁰, there may exist a similar effect in minor ABO-incompatible HTRs. This would mean that hemolysis only occurs when a high enough dose of minor incompatible blood products containing anti-A and anti-B to activate complement is administered to a recipient with an over-activated reticuloendothelial system.

Focusing on antibody load, rather than antibody titer, can have several advantages, especially considering specific clinical scenarios: there is an increasing use of group A plasma as well as group O whole blood for resuscitation in trauma cases due to shortages of blood products such as AB plasma. Many reports have already demonstrated the safety of type A plasma for resuscitation purposes^71–74. In scenarios involving the transfusion of multiple units of group O apheresis platelet units, tracking the cumulative antibody load is critical. It allows for a more accurate evaluation of the overall risk of HTRs as multiple units are administered. This may go against current practice in some blood establishments where a single dilution is used to determine a safe antibody level. However, in multiply transfused patients, minor incompatibility poses an added clinical risk of an HTR, if an arbitrary titer cutoff is applied. Furthermore, it has been shown in previous studies that antibody titer does not vary much over the course of the donor’s lifetime, therefore a one-time titer can be safely used to calculate antibody load for each blood product, and it would be a good predictor for each successive donation⁷⁵.

The current resurgence in the use of whole blood for resuscitation purposes in the emergency setting makes our study even more relevant. In 2011, the Trauma Hemostasis and Oxygenation Research (THOR) Network, an international group of transfusion medicine and emergency medicine specialists as well as researchers, advocated for the use of Low Titer group O Whole Blood (LTOWB) instead of the current standard of component therapy⁷⁶. Various programs worldwide have adopted the usage of LTOWB, including the Norwegian Naval Special Operation Commandos and the U.S. Army’s Armed Services Blood Program⁷⁷, and cold-stored LTOWB is now an FDA-approved product in the U.S. These examples clearly demonstrate a possible resurgence of the use of whole blood and, as LTOWB contains both anti-A and anti-B, it follows that using an antibody load as cutoff might be of better clinical value than simply a titer, as patients requiring massive transfusions often require several products, and the additive effect is not something that can be measured with a qualitative titer alone.

For each blood bank considering changing current practice, it is essential that the determination of a safe cutoff is based on a balance between the risk of causing an HTR in the recipient and the availability of blood products of appropriate blood group.

The main limitation of this study has been that some of the data from case reports are very old, and some studies predate the IAT, and thus the determination of the IgG titer. Titers obtained sixty years ago might differ to those taken on the same sample today. It is difficult to compare titers between laboratories as there are different methods, reagents and local guidelines on titer testing. However, this limitation also applies to other papers that are trying to determine a safe cutoff titer^2–4,60.

Another limitation to this study is the unknown factor of not knowing all HTRs caused by minor incompatible plasma worldwide for the last several decades as not all cases have been reported. Moreover, we don’t know the denominator, i.e. the total number of transfusions during the period of observations.

While we have found and included publications where components with high antibody content did not cause HTRs, there is certainly a risk of publication bias for reports on HTRs. Routine transfusions where patients do not react to components despite a high content of incompatible antibodies are less likely to be published, limiting our ability to draw conclusions on the risk of minor incompatible transfusions.

CONCLUSIONS

This systematic review is to our knowledge the first to introduce the idea of using arbitrary units to estimate specific ABO antibody content instead of antibody titer to gauge the risk of HTRs. Based on our collected data (Table III), a cutoff of 10,000 AU for IgM and 30,000 AU for IgG would mitigate almost all HTR cases involving ABO incompatible antibodies being transfused.

This novel idea of focusing on antibody content instead of titer can increase the safety associated with transfusion of anti-A/anti-B containing blood products such as plasma or apheresis platelet units, which would in turn may increase blood product availability. While it does not exclude the need to standardize titration across blood banks, it does allow for a more empirical evaluation of a patient’s risk for an HTR.