Rare Blood Group Bank in Transfusion Therapy of Patients with Complex Allo-Immunizations: A Single Veneto Center Experience

Abstract

Background: Today, in Western countries, patients with allo-antibodies to high-frequency antigens or with complex antibody mixtures represent one of the most significant challenges in transfusion medicine. Another important aspect is the prevention of allo-immunization of patients who lack high-frequency antigens. In these conditions, the availability of a bank of a rare red blood cell group, supported by a database of donors subjected to extensive erythrocyte typing (preferably using erythrogenomic study), can constitute a resource of great value. Materials and Methods: Repeat Caucasian blood donors of group A or O, with selected Rh phenotypes (CCDee, ccDEE, ccdee, ccDee), aged under 52 years, were considered for typing. Moreover, we selected all non-Caucasian repeat blood donors for typing. For extended phenotyping and genotyping we adopted commercial methods supplied by Grfols and Werfen, respectively. For cryopreservation, we selected a high glycerol method in −80 °C electric freezer; blood unit processing was performed using a Haemonetics ACP 215 automated cell processor with close circuit devices. Results: We considered the five patients as follows: PA was massively transfused for a road trauma, developed multiple allo-antibodies (anti-D, anti-k), and required compatible blood units for an elective cardiac surgery; PB was a pregnant woman with anti-Coa (a high frequency antigen) monitored during pregnancy and in which it was necessary to proceed with the transfusion of the newborn; PC was a poly-transfused patient with myelo dysplastic syndrome who developed multiple allo-antibodies (anti-k, anti-CW, anti-Lea) and required continuous supportive therapy, including the procurement of compatible units and the implementation of therapeutic actions in an attempt to reduce the transfusion requirement using luspatercept; PD was a patient with sickle cell disease. They had a Fy (null) genotype, making it very difficult to find compatible units; and PE was interesting for the complexity of the immunohematological and erythrogenomic study performed to characterize a recipient with a rare phenotype and thus allow the transfusion of compatible units, preventing allo-immunization. Discussion: In this report, we have maintained a narrative approach. Starting with five patients representing as many clinical situations as possible, we have illustrated the approach followed for the immune-hematological study and the choices made to try to guarantee effective and safe transfusion therapy.

1. Introduction

Erythrocyte immunohematology, a branch of laboratory medicine which studies the antigens expressed on the surface of erythrocytes and the related antibodies, was born in 1900 in Vienna with the description of the ABO system by Landsteiner and Wiener. Subsequently, our knowledge in the field of immunohematology has continuously expanded and, currently, the International Society of Blood Transfusion (ISBT) recognizes over 360 red cell antigens classified into 47 blood group systems genetically determined by 52 genes and over 1500 alleles, and three “collections” or “series”. The 200 “series” are designed to group antigens which are biochemically or serologically similar where the genetic basis has not yet been discovered, the 700 “series” are composed of low-frequency antigens (prevalence < 1% in the considered population), and the 900 “series” are composed of high-frequency antigens (prevalence > 90% in the considered population) [1,2,3,4,5]. Currently, a blood group is considered as “rare” if observed in less than 0.1% of the considered population. Some blood groups were considered universally rare, such as the hh genotype (phenotype Bombay), Rh null phenotype, Kell null phenotype, Yt(a−), Vel−, or Kp(b−) [6,7,8]. Conversely, some blood antigens have an unequal distribution in different ethnic groups. For example, RhD− or Fy(a−) may be found only in 0.3% people in Southeast Asia while are more common Caucasian populations; on the other hand, Fy(a−b−) occurs in about two thirds of people of African origin but are extremely rare in Caucasians. So, as general principle, the concept of rare blood should be referred to the reference population [9,10]. Allo-immunization versus red cell antigens may develop after transfusion, transplant, or pregnancy. Transfusion-dependent subjects such as patients with hemoglobin disorders (i.e., thalassemia, HbS) or patients with myelodysplastic syndrome are considered at high risk of developing antibodies against multiple blood group antigens, making it difficult to find compatible blood units, that is, those lacking the antigens to which the patients have been immunized. The process of blind screening multiple blood units to find compatible ones is tedious and time consuming [11,12]. Therefore, programs are multiplying at a national or local level to organize and maintain rare blood donor programs to respond adequately and rapidly to the transfusion requirements of patients with multiple allo-immunizations or with antibodies against high frequency antigens [13,14].

In this paper, starting from the description of some clinical cases that have engaged us during years 2023–2024, we are going to briefly describe the structure and activity of the rare blood group bank of Mestre.

2. Materials and Methods

Study Location and Blood Donors Selection Criteria: This study was performed by the Department of Transfusion Medicine (DIMT) of the Venice prefecture (Northeast Italy). About 35,000 repeat, unpaid blood donors belong to our DIMT. Caucasian blood donors of group A or O, with selected Rh phenotypes (CCDee, ccDEE, ccdee, ccDee), aged under 52 years, were considered for typing. Moreover, we selected all non-Caucasian repeat blood donors for typing [15,16]. Further selection criteria are reported in

Table 1. Blood donors classified as “rare” in our reference Caucasian population.

Oh	S-s-	Fy (a−b−)
CDE/CDE	Lu (a+b−)	Jk (a−b−)
Cde/Cde	Lu (a−b−)	Di (b)−
cdE/cdE	Kp (a+b−)	Co (a−b+)
-D-/-D-	Kp (a−b−)	Co (a−b−)
Rhnull (---/---)	Js (a+b−)	KK

. All subjects gave written informed consent to enroll in this study; the study was performed in accordance with Helsinki declaration. Ethical review and approval were waived for this study because, in Italy, formal ethical committee approval is not required for reports describing clinical cases from routine clinical practice.

Table 1 shows the phenotypes considered rare in a population of Caucasian blood donors because of the absence of high frequency antigens or the combination of multiple genotypes. Repeat blood donors of group O or A aged under 52 years were usually considered. Moreover, all repeat donors of non-Caucasian ethnicity were selected for typing. A third group of subjects who, if available and suitable, could be selected for typing are the relatives of patients with rare blood group red cells.

Blood Typing: Our transfusion medicine has implemented a second-level immunohematology laboratory that has routine access to three different platforms for performing serological tests as follows: test tubes, using a semi-automated in-house method; test cards, using a commercial automated method (Grifols Italia Spa, Milano, Italy); a solid-phase test, using a commercial automated method (Werfen Italia SPA, Milano, Italy). The extended phenotyping protocol adopted in our immunohematology laboratory includes the study of the following erythrocyte blood group antigens: ABO, C, E, c, e, K, k, Kpa,b, Fya,b, Jka,b, M, N, S, s, Lua,b, P1, Xga, Lea,b, using the gel cards method and high productivity automated analyzers supplied by Grifols [17]. Our extended genotyping protocol includes the study of gene coding for the following erythrocyte blood group antigens: C, E, c, e, K, k, Kpa,b, Jsa,b, Fya,b, Jka,b, M, N, S, s, Lua,b, Dia,b, Doa,b, Hy, Joa, Coa,b, Sc1,2, LWa,b, using a commercial automated method (Bead-Chip, Werfen, Barcelona, Spain [18,19]. All methods have been validated and the quality of the results produced is guaranteed by the constant execution of internal quality controls and by participation in external quality verification programs.

Blood Processing: Leukodeplete packed red blood cell units (PRBCu) have been prepared from whole blood using standard blood bank techniques (hematocrit 70% ± 5). PRBCu (volume 200 ± 20 mL) preserved in CPD-A1 (citrate/phosphate buffer containing glucose and adenine) can be stored at 4 °C for up to 6 days before freezing. For frozen PRBCu storage, we selected a high glycerol method in −80 °C electric freezer. RBCu automated processing was performed using the Haemonetics ACP 215 [20,21]. Instrumentation and devices have been qualified, and blood products have been subjected to quality controls to verify the compliance with the required characteristics [22].

The Rare Blood Bank: The rare blood group bank of the Venice district was authorized with resolution of the Veneto Region 3929/1998. Its activity began in 1999 [23], and in 2015, it acquired a high productivity genotyping system. The stated objectives were to reach 10,000 typed blood donors and 500 units of cryopreserved rare blood group units. Actually, 8821 repeat blood donors were subjected to genotyping, of which 5625 were also subjected to extended phenotyping according to the protocols adopted in our transfusion medicine. Of these 8821 repeat blood donors, only 326 (3.7%) were of non-Caucasian ethnicity. Obviously not all PRBCu obtained from typed donors are cryopreserved; as of 31 December 2024, we have 318 frozen blood units in storage.

3. Results

Patient A: PA, a 57-year-old non-Caucasian male, O RHD−, ccee KK (genotyping is reported in

Table 2. Serotyping and genotyping of the five patients described.

	Antigen	Patient A	Patient B	Patient C	Patient D	Patient E
ABO		O	A	O	O	A
RHD	D	−	+	+	+	+
RHCE	C	−	+	+	+	+
	c	+	+	+	+	+
	CW	−	−	−	−	−
	E	−	−	−	−	−
	e	+	+	+	+	+
Kell	K	+	−	−	−	−
	k	−	+	+	+	+
	Kpa	−	−	−	−	−
	Kpb	+	+	+	+	+
	Jsa	−	−	−	−	−
	Jsb	+	+	+	+	+
Kidd	Jka	−	−	−	+	−
	Jkb	+	+	+	−	+
Duffy	Fya	−	−	−	−	−
	FYb	+	+	+	−	+
MNSs	M	+	+	+	+	+
	N	+	+	+	+	+
	S	−	−	−	+	−
	s	+	+	+	+	+
Lutheran	Lua	−	−	−	−	−
	Lub	+	+	+	+	+
Diego	Dia	−	−	−	−	−
	Dib	+	+	+	+	+
Colton	Coa	+	−	+	+	+
	Cob	−	+	−	+	−
Dombrock	Doa	+	+	+	+	+
	Dob	+	−	+	−	−
	Joa	+	+	+	+	+
	Hy	+	+	+	+	+
Scianna	Sc1	+	+	+	+	+
	Sc2	−	−	−	−	−
Lewis	Lea	−	−	−	−	−
	Leb	+	+	−	+	+
Key Findings		Anti-D Anti-k	Anti-Coa	Anti-k Anti-CW Anti-Lea	Anti-Fya Anti Fyb Anti-Jka	RHD RHCE gene fusion

), Direct Antiglobulin Test (DAT) negative, Indirect Antiglobulin Test (IAT) positive for an antibody mixture composed of the following: a first IgG antibody with anti-D specificity (titer 1/128) and a second IgG antibody with anti-k specificity (titer 1/64). The patient’s transfusion history reports, some years ago, a road trauma that required the activation of a massive transfusion protocol with the infusion of twenty-seven units of PRBCu, without the possibility of matching for RhD, RhCE, and Kell antigens [24]. The patient then developed an immune response against the D antigen and the Cellano antigen. Anti-D and anti-K are considered clinically significant because they are able to cause post-transfusion hemolytic reaction (PTHR) and fetal and neonatal hemolytic disease (FNHD); therefore, the transfusion of RhD− and K− units was mandatory [25]. The request for five PRBCs came from another hospital where the patient had to undergo cardiac surgery. In Caucasian blood donors, the frequency of O RhD− is 0.09 and frequency of K− is 0.002, the combined frequency of the two genotypes is therefore about 0.000144 [25,26,27]. In our inventory of cryopreserved red blood cells, we had some O RhD− ccee KK units, two of which, on the morning of surgery, were thawed at our blood establishment and transported at +4 °C to the requesting facility. The cross match of the units was carried out in the destination hospital; both units were transfused without any adverse effects, in the mid-term follow-up, no development of further antibodies was reported.

In this table, the results of serotyping and genotyping performed in the five patients considered in this study are reported.

RhD is the most relevant red blood antigen after the A and B antigens. About 80% of Caucasian are RhD+. It is estimated that 30% to 85% of RhD−subjects receiving RhD+ RBC transfusion will develop an anti-D antibody. Usually, anti-D antibodies are “immune” and not “natural” antibodies which means that a subject lacking the RhD antigen does not regularly present circulating anti-D; so, the formation of anti-D results from immunization after exposure to RBC presenting the RhD antigen, attributable to previous transfusions or pregnancies [25,26,27]. Most anti-D antibodies are IgG, reacting poorly in saline solutions and better in high-protein or enzyme test systems; however, the best method to detect anti-D involves the use of Coombs serum tests. Since anti-D usually does not bind complement, PTHR involving anti-D causes extra vascular hemolysis instead of intravascular hemolysis [25,26,27]. Anti-D fetal maternal immunization was the major cause of HDFN until the introduction of post-natal immune prophylaxis in 1970 that reduced the incidence of maternal RhD allo-immunization from 14% to 1–2%. Subsequently, antenatal immune prophylaxis was also started, reducing RhD allo-immunization to 0.1% in the Western world [28,29,30].

The K antigen is considered quite immunogenic but only 1/500 Caucasian subjects are Cellano-; thus, the detection of an anti-k is relatively infrequent [23,25]. Anti-k antibodies are usually IgG, arising after exposure to RBC, for example, due to blood transfusions or pregnancy. Usually, they are reactive in anti-globulin-based assays [28,31]. Anti-k can cause both immediate and delayed PTHR and have been associated with cases of mild to severe HDFN [31].

Patient B: PB, a 38-year-old Caucasian female, come to our observation in September 2023 at 28 weeks of pregnancy. PB was A RhD+, Ccee kk, DAT-IAT+, pan-reactive (titer 1/64). The patient’s history reported no blood transfusion, a previous pregnancy in 2020, and a miscarriage in January 2021. We performed extensive phenotyping and genotyping of PB (see Table 2) and her partner, finding that PB presented the rare Coa-Cob+ genotype (reported in less than 0.3% of the Caucasian population) while the husband presented the common Coa+b− genotype [25,26,27]. Therefore, suspecting the presence of an allo-antibody with anti-Coa specificity, we sent a sample to the Immunohematology Reference Laboratory of Policlinico in Milan, which confirmed the presence of an anti-Coa antibody. Usually, anti-Coa antibodies are IgG that react with Coombs serum, with enzyme treatment of enhancing red cells reactions. Anti-Coa are considered clinically significant because of their association with PTH and HDFN [30,32]. PB was monitored during pregnancy with an IAT (with determination of the antibody titer) every two weeks. Moreover, echographic evaluation of peak flow velocity of median cerebral artery, with fetal hemoglobin estimation was performed. The anti-Coa antibody titer progressively increased to 1/256, delivery took place on 27 October 2023, gestational age 34 weeks, by elective cesarean section. Transfusion medicine had been updated about the evolution of the clinical situation; therefore, after consulting our database of typed donors, two subjects were called for bloodletting in order to have available fresh Coa- blood units. Timely steps were taken to prepare irradiated pediatric aliquots so that transfusion support was readily available in case of need. The baby was AB RhD+ DAT+ (+++−) with anemia and jaundice. After admission to the Neonatal Intensive Care Units, pan phototherapy was started in association with hydration and intravenous single dose (1 g/Kg) IG administration. The baby required transfusion of two pediatric units but not exchange transfusion. The baby was discharged on 9 November. One year after birth, the baby had regular psycho-somatic growth. An interesting aspect is that PB, after the suspension period from donations, which according to the Italian national regulation is 12 months [33,34], joined our rare blood group banking program and in December 2024 made the first donation of whole blood. The concentrated red blood cells were cryopreserved.

Patient C: PC, a 73-year-old Caucasian female, O RhD+ Ccee KK, affected by myelodysplastic syndrome, unresponsive to treatment with IV iron and erythropoietin, transfusion-dependent. The patient was sent from another hospital because she had multiple anti-erythrocyte antibodies with specificity anti-k (titer 1/16), anti-CW (titer 1/4), and anti-Lea (titer 1/4), which made it difficult to find compatible units.

Among Caucasians, the k antigen is considered a high frequency antigen because it is present in about 99.8% of cases and only 0.2% are k-. The characteristics of anti-k antibodies have already been discussed in the presentation of the first clinical case. The C Willis (CW) antigen is quite uncommon because it is only present in 1–2% of the population. In addition, most subjects of CW+ are also C+. Anti-CW antibody was first described in 1946. It often occurs naturally as well as in combination with other antibodies to red blood cells antigens. Anti-CW is usually of the IgG type and has the potential to cause mild to severe immediate or delayed PTHR and mild to moderate FNHD [35,36]. The Lewis antigens are plasma proteins that readily adsorb to and elute from red cell membranes. Lewis antibodies occur almost exclusively in Le(a−b−) individuals, often without known red cell stimulus, most Lewis antibodies agglutinate in saline assays but some anti-Lea can be detected in the antiglobulin phase of testing. Transfused red cells shed their Lewis antigens and assume the Lewis phenotype of the recipient within a few days of entering the circulation; therefore, anti-Lea are seldom associated with PTHR. Lewis antibodies are usually IgM and do not cross the placenta; furthermore, Lewis antigens are poorly developed at birth, so anti-Lea antibodies are not associated with HDFN [37]. PC was taken into care by our facility and, as usual for recently transfused patients, we proceeded with an extended genotyping whose results are reported in Table 2. As regards the treatment of anemia, we followed a dual approach as follows: first, transfusion support was provided using units KK and CW− PRBCu using both cryopreserved rare group red blood cells stocks and, preferably, by calling typed donors. Moreover, we successfully used Reblozyl (luspatercept) to reduce the patient’s transfusion requirement [38,39,40]. In a mid-term follow-up, no development of further antibodies was reported.

Patient D: PD, a 25-year-old female from Central Africa, affected by sickle cell anemia, came to our attention in April 2023, following a night-time access to the emergency rooms (ER) of dell’Angelo Hospital because the painful sickle cell crisis was immediately treated with intravenous opioids and hydration. PD did not have any documentation relating to previous transfusion therapy but reported having received several transfusions in various healthcare facilities in Italy and abroad. The patient reports a previous spontaneous miscarriage. PD was O RhD+ Ccee, kk IAT+ DAT+. At presentation, a mixture of anti-erythrocyte and allo-antibodies were detected as follows: anti-Fya (titer 1/2) and anti-Fyb (titer 1/2). The Rare Blood Bank of Mestre had four frozen Duffy null units and two eligible rare donors on its database, but, for logistical reasons, it is not possible to proceed with defrosting at night. Anyway, due to symptomatic anemia, a nightly urgent transfusion with two incompatible (Fya+,Fyb+) was imposed by the clinician in charge of ER, and no post-transfusion reactions were observed. The next morning, two Fya−b− units were thawed and transfused. In any case, it was decided to contact the Ospedale Maggiore Policlinico in Milan, home of the main Italian rare blood group bank, to support PD transfusion therapy [41,42].

Duffy glycoprotein is the RBC receptor for Plasmodium vivax, so RBC lacking Duffy antigens are relatively resistant to infestation by P. vivax. This has influenced the Duffy blood group variation observed in populations where malaria is common [43,44]. The Fy(a−b−) “null” phenotype is found in 68% of Blacks but is extremely rare in Caucasians, Chinese, and Japanese. The allele involved in the majority of these cases is FY*02N.01 with a mutation of the nucleotide −67t > c^ and causes the absence of Duffy glycoproteins on red blood cells [24,27]. In a patient previously transfused, our standard approach foresees an extensive genotyping whose results are reported in Table 2. Anti-Fya is a common antibody while anti-Fyb is 20 times less common, these antibodies are usually IgG and react best at the antiglobulin phase. Some examples of anti-Fya and anti-Fyb show dosage, reacting more strongly with homozygous than heterozygous RBCs. Duffy antigens are cleaved by protease, so anti-Fya and anti-Fyb do not react with enzyme-treated RBCs. Anti-Fya and anti-Fyb have been associated with mild acute and delayed PTHD and with light to mild HDFN [24,25].

PD presented three further hemolytic crises in May, July, and August and, in each case, two Fya−b− PRCu were transfused. No adverse effects were observed although a modest increase in antibodies titer (1/8) was observed [45]. From April to August 2023, PD received two Fya+b+ and 8 Fya−b− units. In August 2023, a third antibody with anti-Jkb specificity appeared (titer 1/2). This new immunization made it even more complicated to find compatible units. In fact, in subjects of African ethnicity, if the frequency of the Fya−b− phenotype is 0.68, the frequency of Jkb- subjects is 0.51; therefore, the possibility of finding a compatible donor drops to 0.35. Anti-Jka is more frequent than anti-Jkb, but are both quite uncommon, these antibodies are usually IgG and react best with the antiglobulin phase. Many anti-Jka and anti-Jkb react more strongly with RBCs that carry a double dose of the respective antigen (dosage effect). The main feature of the anti-Jka and anti-Jkb is the rapid titer decrease in vivo. Kidd antibodies are a common cause of PTHR, especially of the delayed type; on the other hand, Kidd antibodies are only rarely associated with severe cases of HDFN [46,47,48]. Patient PD subsequently moved abroad and is no longer followed at our facility.

Patient E: PE, a Caucasian male aged 72, affected by acute myeloid leukemia (AML), was treated with azacytidine and developed severe pancytopenia, requiring constant transfusion support with platelet concentrates and PRBCu. The phenotype of PE was A RhD− CCee kk. IAT−, DAT+ (++−−), Dweak+ (++−−), possibly due to DAT+. Phenotype dCe/dCe is extremely rare (<0.1%) among Caucasians, so the patient was genotyped with the following results: A, RHD+ (no D variant were observed), RHCE (C+c−E−e+), Kell (K−k+), Co (a+b−), Di (a−b+), Do (a−b+), Fy (a−b+), Jk (a+b+), Js (a−b+), Kp (a−b+), Lu (a−b+), Lw (a+b−), M+, n−, S+, s−, Sc (1+2−). A discrepancy between phenotype and genotype was observed. Therefore, we shipped a sample to the New York Blood Center to proceed with the sequencing of the RHD and RHCE genes which was carried out both with the PCR-multiplex method for exons 4 and 7 of the RHD gene. RHD* pseudogenes were searched for, and genotyping was performed for the RHC and RHc genes; furthermore, the sequencing of the RH region was carried out using a NGS-based assay. Sequencing confirmed the presence of the RHD gene (exon 4,7), the absence of RHD* pseudogenes, RH*C/C. The genotyping results do not predict an altered D phenotype. His phenotype may be caused by a genetic alteration not interrogated by NYBC assays but, unusually, the sample appears to have one RHD replaced by that of RHCE*Ce. This may be the Ceppellini effect, given to homozygous RHCE*Ce and the possible hybrid RHD-Ce(1–9) allele, also predicted to cause a D-phenotype [49,50]. Moreover, we performed an adsorption–elution test on a fresh blood sample that indicated an antigen titer of 1, which denotes a very low RhD antigen density. Alternatively, this finding could suggest that one RHD gene was replaced by that of RHCE*Ce gene, RhCE protein can express RhD epitopes detected by some monoclonal anti-D. There are rare individuals who possess these unusual proteins and, when typed with anti-D, will show a very weak positive reactivity even though the D epitope is on the RhCE protein [51,52]. In any case, we decided to continue as far as possible and transfuse PE with units of PRBC RhD-CCee. To date, the patient, received eleven PRBCu units and is still negative for anti-erythrocyte antibodies.

4. Discussion

In clinical practice, patients who pose the greatest problems in obtaining compatible PRBCu units are those who present allo-antibodies to high-frequency antigens or who present complex antibody mixtures. Another important aspect is the prevention of allo-immunization of patients who lack high-frequency antigens. In these conditions, the availability of a rare red blood cell group bank, supported by a database of donors subjected to extensive erythrocyte typing (preferably by erythrogenomic study) can constitute a resource of great value [53,54,55].

The Mestre rare blood cell bank began its activity in 1999. Due to organizational difficulties, activity slowed down in 2008, only to resume in 2015 with the acquisition of a high-throughput system for genotyping and an automated system for PRCu glycerolization/deglycerolization treatments. The choice to cryopreserve the PRBCu in high glycerol at −80 °C in an electric freezer rather than using liquid nitrogen was dictated by logistical considerations, as the processing and cryopreservation area is entirely dedicated to blood stem cells. Regarding the shelf life of frozen red blood cells, the EDQM provides for 30 years [22], while Italian national regulations provides for 10 years, with the possibility of extending to 30 years for very rare groups [33,34]. At the Department of Transfusion Medicine in Venice, we apply the principle of deferred first donation: prospective donors are assessed and, if eligible, invited to make their first donation after four weeks. Both during the suitability assessment and at the first donation, we determine the donor’s ABO group, RhD, RhCE, and Kk [33,34]. Therefore, at the time of the second donation (when the donor is considered a repeat donor), we have the necessary data to select the subjects for extended phenotyping and genotyping. The selection criteria are those recommended by ISBT [56,57,58,59]. The number of subjects selected is limited by the available budget, which allows us to perform extended typing on approximately 1200 subjects per year, including patients.

Usually, protocols for massive transfusion of patients of an unknown blood group provide for the initial use of emergency group O ccdee kk RBC; if the blood group is known, or once determined, ABO- and D-compatible units should be used. As far as possible, RhD-RBC should be used for RhD− patients, especially females with child-bearing potential. With RhD− males and older females, a decision may be made to use RhD+ RBC considering blood’s availability and the need to maintain adequate stocks of RhD− PRBCu [60]. In the first case presented, patient PA underwent a massive transfusion following a severe road accident and it was necessary to transfuse him massively without the possibility of respecting the RhD, RhCE, and Kk compatibility. As a consequence, the patient developed, an anti-D but not anti-C and/or anti-E. Less predictable was the production of anti-k allo-antibodies which are much less common and cause greater problems in finding compatible units [61].

Pregnancy remains an event potentially capable of immunizing the mother against paternally inherited antigens present on fetal RBC. HDFN are usually classified into three categories, such as disease caused by anti-D alone or in combination or other allo-antibodies against antigen of RH system; disease caused by other allo-antibodies (more frequent are anti-c and anti-K); disease caused by ABO incompatibility [24,25,28,29]. Diagnosis and management of anti-Coa allo-immunization during pregnancy is a rare and challenging condition and only few cases have been reported in the literature, mainly in the 1970. Moreover, diagnosis and management has changed in the meantime [62]. In order to closely monitor a fetus at risk of developing HDFN complications, a pregnancy follow-up with repeated Doppler flow measurements of the peak systolic velocity of the middle cerebral artery is recommended. On the other hand, antibody titration seems to be of little help in defining the risk of developing fetal anemia, due to the lack of a correlation between the observed titer increase and the onset of clinical manifestation, and to the fact that different analytical methods may result in different titration values [63]. The newborn required admission in an intensive neonatal care unit and required blood transfusion. Transfusion therapy in these patients should be quite problematic due to the difficulty of finding Coa- red blood cells but, in this case, PRBC units were granted by our rare blood bank with the aid of the transfusion network of the Veneto Region.

Myelodysplastic syndromes (MDS) are myeloid clonal diseases characterized by ineffective hemaptopoiesis with peripheral cytopenias and an increased risk of leukemic transformation. Transfusion remains a cornerstone of management for MDS patients. As a matter of fact, up to 90% of patients require RBC transfusions during the course of their disease and 30–45% become dependent on RBC transfusions. In MDS patients, the cumulative incidence of allo-immunization was 11%. Allo-antibodies were most commonly directed against antigens in the Rh (54%) and Kell (24%) systems. Multiple allo-antibodies were present in 49% of allo-immunized patients. In allo-immunized patients, transfusion requirements are higher than in patients without allo-antibodies; furthermore, a significantly higher proportion of allo-immunized patients than non-allo-immunized patients had detectable auto-antibodies [64,65]. Patient PC was sent to our facility from another hospital due to the presence of an antibody mixture (anti-k+anti CW anti-Lea) which made it extremely difficult to find compatible units. Due to their limited clinical relevance, it was deemed possible to ignore the antibodies with anti-Lea specificity, and we concentrated on ensuring the necessary transfusion support with KK CW- units both by recalling typed donors and by using cryopreserved PRBCu [24,25,34,35,36]. Considering the experience gained in our facility in the treatment of patients with transfusion-dependent beta-thalassemia using Luspatercept, we have successfully used Rezobil to reduce the patient’s transfusion requirement [66,67].

In Western countries, the number of people affected by sickle cell disease (SCD) is constantly increasing because of migratory flows from Africa and the Middle East. However, selecting compatible units can be challenging, especially in the case of patients of sub-Saharan (SSA) descent. In these populations, malaria is a major pressure, favoring the selection of hemoglobin disorders such as sickle cell trait and rare blood groups (such as Fya−b−), conferring partial protection against Plasmodium infection [43,44]. In Italy, patients are commonly transfused with blood collected from Caucasian donors, and therefore a mismatch of red cell antigens occurs very frequently [68]. Moreover, in previously transfused SCD patients, the frequency of allo-immunization can be as high as 70%, predisposing them to PTRH and causing a progressive restriction of compatible donors [69,70]. With the aim to reduce allo-immunization frequency, it is now recommended that patients with SCD (and transfusion-dependent thalassemia) undergo extended RBC serotyping and/or genotyping as soon as possible (optimally, before the first transfusion) and, whenever possible, receive RBC transfusions prophylactically best matched for ABO, Rh (C, E or C/c, E/e), Kell (K/k), Kidd (Jka/Jkb), Duffy (Fya/Fyb), and MNS (S/s). However, to ensure matched red blood cells, a suitable number of donors of African ancestry need to be included in the donor pool. For example, a ccDee phenotype is quite frequent (about 60%) of SCD patients but in less than 3% of blood donors of Italian ancestry. Moreover, the Duffy null phenotype (Fya−/Fyb−) is present in the vast majority of SCD patients but nearly absent among Italian donors [69,70]. In our facility, we have extensively genotyped and serotyped nearly 9000 repeat blood donors, of which only 326 (3.7%) were non-Caucasian. Undoubtedly, cultural and linguistic barriers, as well as a different sensitivity towards the culture of voluntary, anonymous, and unpaid blood donations, play a role in the recruitment of foreigners. For example, Italian legislation requires that the donor be able to understand and to fill in correctly a standardized questionnaire prepared at a national level, which must be exclusively in Italian. A further difficulty for non-Caucasian subjects is traveling to their countries of origin, which often, under Italian national legislation, entails prolonged periods of suspension from donations [32,33]. In the case of PD, with the precious support of the Rare Red Blood Cell Group Bank of Milan, we were able, in election, to ensure the transfusion of PRBCu Fya−b−; however, it was not possible to fully respect the phenotype of the patient who developed a third antibody with anti-Jkb anti-Kkb specificity. The patient subsequently moved abroad and was no longer treated at our facility.

In leukemic patients, the allo-immunization rates vary from 1 to 3% for acute lymphoblastic leukemia to 12–15% for acute myeloid leukemia. In patients with AML, despite the possible problems due to allo-immunization (i.e., post-transfusion reaction and limit in blood product availability), no differences in overall survival were observed between allo-immunized and non-allo-immunized patients [71]. In any case, efforts to reduce RBC allo-immunization either by extended serologic matching or genotyping by PCR should become the future standard of care for patients with a foreseeable long-term transfusion dependency [72,73]. The case of PE seemed particularly interesting to us because, from the observation of an extremely rare CCdee phenotype (<0.1%) in the Caucasian population, we proceeded with in-depth genotyping investigations that excluded the presence of weak RHD or variant RHD. Through genotyping, performed at the New York Blood Center, we demonstrated that PE was one of those rare individuals who possess these unusual proteins and, when typed with anti-D, will show a very weak positive reactivity even though the D epitope is on the RhCE protein [49,50,51,52]. Applying the precautionary principle, PE was always transfused with RhD- RhCe PRBCu [74]. It seems likely that this rare genotype may be associated with the formation of anti-D when RhD+ red blood cells are transfused. Therefore, in this case, the use of rare blood group (CCdee) prevented the allo-immunization of difficult clinical management [75].

Study limitations: The present study has several limitations. First, uncertainties regarding the regular disbursement of dedicated funding have created difficulties in obtaining and maintaining adequate resources, with significant repercussions on the operations of the rare blood group bank. For example, the goal of 10,000 typed donors was only 80% achieved, and the target of 500 cryopreserved rare blood group units was only 64% achieved. Secondly, following the 1998 authorization resolution, no regional indication was followed to centralize the activities of extensive donor typing and cryopreservation of rare blood cells.

However, in our opinion, the main limitation is that our rare blood cell bank, until now, has had only a local role, being part of the Veneto Region’s transfusion network but with little integration at the national level. In our opinion, with the recent implementation of European standards for the substance of human origin (SOHO) [76] at the Italian level, rare blood cell banks will be considered “critical resources,” and this will require the implementation of a network at the national and supranational levels.

5. Conclusions

In this paper, we have presented five interesting cases to describe the therapeutic challenges that must be addressed and resolved to ensure safe and effective transfusion therapy in complex immunohematological cases. The first case involved a patient who was massively transfused following a poly-trauma and who developed multiple allo-antibodies and required the procurement of compatible units for cardiac surgery. The second case involved a pregnant patient who had developed an allo-antibody to a high frequency antigen. The third case involved a patient who was poly-transfused for MDS and who had developed multiple allo-antibodies and required continuous supportive therapy, through the procurement of compatible units and the implementation of therapeutic actions to reduce the transfusion requirement. The fourth case involved a patient with sickle cell disease who presented an Fy(a−b−) phenotype for which it was particularly difficult to find compatible units and for which we had to resort to the support of the rare blood group bank in Milan. The fifth and final case, in our opinion, is interesting for the complexity of the immunohematological and erythrogenomic study performed to characterize a recipient with a rare phenotype and thus allow the transfusion of compatible units, preventing allo-immunization.

In ensuring effective and safe transfusion therapy in these five patients, in our opinion, several factors contributed to this academic field: first of all, the availability of an immunohematology laboratory capable of dealing with even remarkably complex clinical cases. Secondly, the availability at our facility of a cryopreserved rare red blood cell group bank was supported by a database of extensively phenotyped and genotyped donors who were available to be called if necessary. Thirdly, it was crucial to have a close collaboration not only with the Veneto Region transfusion network but also with the rare red blood cell group bank in Milan.

In any case, it was essential to maintain a multidisciplinary approach by constantly interacting with the clinicians who were in charge of the patients.