5.3 Immunogenicity and Antigen Prevalence

5.3 Immunogenicity and Antigen Prevalence

Quick Answer: After ABO, D is the most immunogenic red-cell antigen. The classic ranking is D >> K > c > E > k, Fya, Jka, S. Antigen frequencies differ by ancestry — never use the Caucasian column for every patient. Fy(a−b−) is common in African ancestry because a GATA-1 promoter mutation silences FYB on red cells only (tissues still express Fyb; malaria P. vivax resistance). High-prevalence antigen-negative units are rare and come from rare-donor files. Lewis antibodies are usually not clinically significant — do not reflex Le(a−b−) units.

Immunogenicity is not the same as frequency

Immunogenicity is the probability that an antigen-negative person will form an alloantibody after exposure. Frequency is how often the antigen appears in a population. K is uncommon (~9% of European-ancestry donors) and highly immunogenic. Antigen e is almost universal and weakly immunogenic. Mixing those ideas produces the wrong antibody to expect after a unit of e+ blood. Outline II.C (June 9, 2026) tests both the ranking and the ancestry table.

Classic ranking the BB exam still uses (protein antigens, after ABO incompatibility):

D >> K > c > E > k, Fya, Jka, S (then C, e, and the rest).

Numbers worth an item: historically, about 80% of D− recipients of a full D+ red-cell unit formed anti-D (modern leukoreduced, shared-decision practice is lower, but “D is in a class by itself” is unchanged). K immunizes on the order of 10% of exposed K− recipients — far above Duffy or Kidd. That is why K is in every prenatal panel conversation and why many programs prophylactically match K for girls and women of childbearing potential even before an antibody appears. c and E are the next Rh troublemakers; anti-e and anti-C are less common because almost everyone is e+ and because C is less immunogenic than c. Jka is moderately immunogenic and infamous for disappearing from the screen, then causing a delayed hemolytic or delayed serologic transfusion reaction. S and Fya sit in the same “clinically significant, not D/K-level” band.

ABO remains outside this list because anti-A and anti-B are naturally occurring. You do not need a transfusion to make them. D, K, and Jka require exposure (transfusion or pregnancy), with rare “naturally occurring” exceptions the exam will flag as unusual.

Antigen frequencies you must not flatten into one U.S. number

Approximate percent antigen-positive values the exam expects (adults; rounded textbook figures):

Read the table as a decision tool, not trivia. A D− unit is uncommon in East Asian donors (~1%) and merely uncommon in European donors (~15%). A C− unit is easy to find in African-ancestry inventory and hard to find in East Asian inventory. K− is the default Asian unit and still 91% of European units. Fy(a−b−) is an African-ancestry phenotype, not a typing error. s− (often U-var or U− in African ancestry when S−s−) is a rare-donor problem, not a European-panel problem.

Why Fy(a−b−) exists — GATA-1, tissues, malaria

The common African FY-null RBC phenotype is not a deleted Duffy gene. It is a point mutation in the GATA-1 erythroid promoter of ACKR1 (T>C at the GATA motif, classically on the FYB allele). Erythroid transcription factor GATA-1 can no longer bind, so red cells make no Duffy. Non-erythroid tissues still transcribe FYB, so the person expresses Fyb on endothelium. Consequences the exam loves:

• Red cells type Fy(a−b−).
• The patient can make anti-Fya if exposed to Fya (they have no Fya anywhere).
• The patient usually does not make anti-Fyb, because tissues display Fyb — it is “self.”
• A microarray that reads only the 125G>A Fya/Fyb SNP will call FYB / Fyb+ unless the assay also targets the GATA-1 promoter SNP.
• Duffy on red cells is the receptor for Plasmodium vivax (and P. knowlesi). The silenced erythroid phenotype is the human adaptation that reduces P. vivax invasion. It is not protection against P. falciparum.

Do not issue Fy(b−) blood as a standing order for every Fy(a−b−) African-ancestry patient “just in case.” Match for Fya if they have anti-Fya. Do not invent anti-Fyb they have not made. True FY gene deletion (rare) can make anti-Fy3; the common GATA-1 African phenotype does not behave that way.

High-prevalence antigens and the rare unit

k, Kpb, Jsb, U, Vel, Yta, Coa, Ge, Lub, Kx are high-prevalence: >99% of unselected donors are positive. The antigen-negative unit is rare. You will not find two k− units by screening the 20 O-negative walk-in units in the refrigerator. Those requests go to a rare donor registry, autologous donation, siblings (25% chance of the same recessive-looking null if both parents are heterozygotes), or — when clinically justified — least-incompatible blood plus medical-director approval. Jsb− is more often found in African-ancestry donors; Kpb− is not. Searching the wrong ethnic donor file wastes the night.

Selecting antigen-negative blood and calculating the screen

Once the antibody is identified and is clinically significant (reactive at 37 °C / AHG, associated with HDFN or hemolytic transfusion reactions):

1. Type the patient for the corresponding antigen — they should be negative.
2. Crossmatch antigen-negative units. For multiple antibodies, the unit must be negative for every implicated antigen.
3. Calculate how many units to pull for typing: n / (product of antigen-negative frequencies in the correct population).

Example: African-ancestry patient with anti-E and anti-C. C− ≈ 73% (1 − 0.27), E− ≈ 78% (1 − 0.22). Combined 0.73 × 0.78 ≈ 0.57. Two units: 2 / 0.57 ≈ 4 units to screen. If you wrongly used European frequencies (C− 30%, E− 70% → 0.21), you would tell the floor you need 10 units and delay a surgery that four typed units would have covered.

Give prophylactic extended matching (C, c, E, K, and often Jka, Fya) to sickle-cell and other chronically transfused patients before they make the antibody — that is prevalence plus immunogenicity used as prevention.

Traps that cost points

Using Caucasian frequencies for all patients. An African-ancestry patient needing Fy(a−) blood is easy (only ~10% are Fya+). A European-ancestry patient needing Fy(a−) is not (~34% Fya−). The reverse is true for c− units.

Calling Lewis clinically significant. Anti-Lea and anti-Leb are typically IgM, often naturally occurring, and usually do not react at 37 °C / AHG. They do not cause HDFN (Lewis antigens are adsorbed, poorly developed at birth, and the antibodies do not cross the placenta as IgG). You do not routinely provide Le(a−b−) units. Neutralization with saliva or a 37 °C screen that is negative lets you ignore them. Rare AHG-reactive Lewis antibodies exist; the default exam answer is still “not clinically significant.”

Confusing immunogenicity with prevalence. “Everyone is e+, so anti-e must be common” is backwards. “K is rare, so anti-K must be rare” is also backwards — K is uncommon and very immunogenic, and anti-K is one of the most frequent immune antibodies after ABO and D.

Treating Fy(a−b−) as a Duffy-null person who will routinely make anti-Fy3 or anti-Fyb. The common GATA-1 African phenotype does not.

Screening the refrigerator for k− or U− units. High-prevalence-negative blood is a rare-donor problem.