What is a donor-specific antibody (DSA)? The term explained
A donor-specific antibody (DSA) is an antibody in a transplant recipient that is directed against an HLA molecule the donor carries. In the words of a 2022 review of virtual crossmatch testing (Bhaskaran, Heidt and Muthukumar, Kidney International Reports), when single antigen bead assays "detect the presence of antibodies that are directed against donor HLA, then these are called donor-specific antibodies (DSAs)".
DSA is therefore not a property of the antibody alone. The same antibody is a DSA against one donor and not against another, and deciding which needs two inputs: the recipient's antibody specificities and the donor's HLA typing. This guide explains the term and the role HLA nomenclature plays in it. It does not give clinical advice; what a DSA means for a particular patient is for the transplant team and its laboratory.
The definition in practice
The same review gives a simple example: a patient may have circulating antibodies against HLA-A2, and "if this patient's prospective donor has HLA-A2, then this antibody against HLA-A2 is called a DSA". It also notes that, although the term is used generically, "the term DSA—as detected in most laboratories—implies anti-HLA IgG antibodies".
Definitions are not uniform. The abstract of the Sensitization in Transplantation: Assessment of Risk (STAR) 2017 working group report states that the criteria used to define the presence of a DSA and assign it a level of risk "vary markedly between centers". When you read a DSA result, read the laboratory's own definition with it.
Preformed and de novo DSA
Two qualifiers describe when a DSA is found:
| Term | Usage in the literature |
|---|---|
| Preformed (also "preexisting") | DSA already present before transplantation |
| De novo | DSA that appears after transplantation |
The STAR 2017 abstract speaks of "preexisting (memory) or de novo donor-specific HLA antibodies", and the abstract of the STAR 2022 consensus document on post-transplant assessment of DSA groups antibodies by DSA status "(ie, preformed or de novo)". The distinction is about timing relative to the transplant, not about a different kind of antibody.
How a DSA is identified
The usual route, as the 2022 review describes it, combines two independent tests:
- Antibody testing. A solid-phase assay, in which one reactant "is bound to a solid surface such as a microsphere", detects which HLA molecules the recipient's serum reacts with. In single antigen bead (SAB) assays, each bead is "coated with a single purified HLA protein at an allelic level", for example
B*44:02andB*44:03on separate beads. - Donor HLA typing. The donor's alleles at the relevant loci.
The antibody specificities are then compared with the donor's typing. Any specificity that matches something the donor carries is a DSA. The virtual crossmatch formalises this comparison: the review quotes the American Society for Histocompatibility and Immunogenetics (ASHI) definition of it as "an assessment of immunologic compatibility based on patient's alloantibody profile compared with donor's histocompatibility antigens", and notes that it "does not involve mixing of serum and cells", unlike a physical crossmatch.
Why donor typing resolution decides the answer
Because SAB beads can carry allele-level proteins, an antibody can react with one allele and not with a close relative. The 2022 review's example: if a patient has antibody against B*44:02 but not B*44:03, and the donor is typed only at the first field, as B44, then "it is not possible to determine whether the antibody against B*44:02 is donor specific or not".
This is an HLA nomenclature problem as much as an antibody one. B*44:02 and B*44:03 differ in the second field, which by the WHO naming rules means their encoded proteins differ in amino acid sequence. A first-field typing, B44 or B*44, cannot tell them apart.
The review says the virtual crossmatch "relies on the precise and complete donor HLA typing". Two consequences follow for anyone handling the data:
- Resolution. An antibody specificity written at allele level can only be compared with donor typing at the same or finer resolution. A two-field name fixes the protein; a first-field name does not.
- Loci. DQ and DP molecules are built from two chains. The review notes that HLA-DQ has an α chain encoded by HLA-DQA1 and a β chain encoded by HLA-DQB1, and HLA-DP likewise by DPA1 and DPB1. Comparing antibodies against these molecules needs typing for both chains' loci.
The typing resolution guide covers what each level of a typing result can and cannot distinguish.
MFI is not a titre
SAB results are reported as mean or median fluorescence intensity (MFI). The 2022 review describes MFI as "a semiquantitative assessment that does not reflect accurately the concentration or titer of the antibody", and states that "cutoff values, thus, are different among laboratories and have not been standardized". This guide does not give thresholds. Any cut-off belongs to the laboratory that set it and should be read with that laboratory's definition.
The review also notes that "the failure to detect an antibody on a SAB assay is not a proof of absence of that antibody in circulation", and that bead sets cover the most prevalent HLA in a predominantly Caucasian population. An allele with no bead cannot be tested directly.
Related terms
| Term | Meaning, as used in the 2022 review |
|---|---|
| Virtual crossmatch | Comparing the recipient's antibody profile with the donor's HLA, without mixing serum and cells |
| Unacceptable antigens | HLA the patient has antibody to, above a laboratory threshold or shown to be cytotoxic, reported as unacceptable for the patient; which HLA to list "is left to the discretion of individual transplant program" |
| Calculated PRA (cPRA) | "a value that is based on the unacceptable antigens" |
Getting the names right
A DSA assignment is a comparison of names: antibody specificities on one side and donor typing on the other. Both have to use current, valid names from the same IPD-IMGT/HLA release before the comparison means anything. A legacy spelling, a deleted name or a G group on one side can make two descriptions of the same protein look different, or two different proteins look the same.
GET /v1/allele/{name} reports, for a name in the pinned release, whether it is assigned, its two-field form and its G and P groups. HLA-Verify checks names; it does not interpret antibody results and does not assign DSA.
Frequently asked questions
What does donor-specific antibody mean?
An antibody in the recipient directed against an HLA molecule the donor carries. It is identified by comparing the recipient's antibody specificities with the donor's HLA typing.
What is the difference between preformed and de novo DSA?
Timing. Preformed DSA is present before transplantation; de novo DSA appears afterwards.
Why does donor typing resolution matter for DSA?
Bead assays can show antibody to one allele and not a close relative, such as B*44:02 but not B*44:03. A donor typed only as B44 cannot be compared with that result.
Is MFI the same as antibody titre?
No. The 2022 review describes MFI as semiquantitative and not an accurate reflection of antibody concentration or titre, and notes that cut-offs differ between laboratories.
Sources
- https://doi.org/10.1016/j.ekir.2022.03.006
- https://doi.org/10.1111/ajt.14752
- https://doi.org/10.1016/j.ajt.2022.11.013
- https://hla.alleles.org/pages/nomenclature/naming_alleles/
HLA-Verify is a research-and-evaluation tool for nomenclature and reference-release validation. It is not a medical device and this guide is not clinical advice. Send allele names, never patient identifiers.