A generalized HLA prediction model for related donor matches

Area of Science:

• Immunogenetics
• Population Genetics
• Medical Statistics

Background:

• Finding Human Leukocyte Antigen (HLA) matches is crucial for transplantation.
• Sibling searches are a primary method, but often unsuccessful.
• Alternative strategies are needed to increase HLA matching success rates.

Purpose of the Study:

• To develop a probabilistic model for predicting HLA matches within extended family.
• To evaluate the success rate of extended family searches for HLA matching.
• To assess the impact of ethnic background and family size on HLA matching probabilities.

Main Methods:

• Developed a generalized probabilistic model using family pedigree data.
• Incorporated haplotype frequency data specific to various ethnic groups.
• Calculated likelihood of HLA matches among descendants of grandparents (excluding parents/siblings).

Main Results:

• Extended family searches can achieve success rates comparable to or exceeding the 0.25 rate of single sibling searches.
• Success is higher when patients and family members share common ethnic haplotypes.
• Families with >2 children/nuclear family show significantly higher success rates.

Conclusions:

• Extended family searches offer a viable strategy for HLA matching, particularly for patients with common haplotypes.
• The model provides valuable insights for optimizing donor searches in diverse populations.
• A significant percentage of the population in countries like Japan, the US, and UK could benefit from these extended family searches.