OBJECTIVE: Preimplantation Genetic Diagnosis (PGD) of single gene disorders, combined with HLA matching, has recently emerged as a tool, for couples at risk of passing on a genetic disease, to select embryos both free of the disease and of a compatible Human Leukocyte Antigen (HLA) tissue type with an existing affected child. In this study, our aim was to optimize a flexible preimplatation HLA matching protocol appropriate for the analysis of a wide spectrum of possible HLA genotypes, precluding the design of case specific protocols each time. DESIGN: An indirect single-cell HLA typing protocol, based on a multiplex fluorescent polymerase chain reaction (PCR) of microsatellite markers located in HLA complex, was designed. The strategy was applied on 23 PGD cycles. MATERIALS AND METHODS: External and internal primers for amplification of polymorphic STR markers scattered throughout the HLA complex were designed. The primer sequences of the individual STRs were chosen to work in a multiplex fluorescent PCR format, in combination with the other markers and the primers used to identify the disease. A panel of 29 different STRs were studied during the set-up phase to evaluate their informativity for HLA matching in family members (father, mother, and affected child). The PGD strategy involved a multiplex external/nested PCR amplification of the causative mutations, STR markers linked to these regions for ADO detection, and a range of 10-15 informative HLA markers simultaneously from biopsied blastomeres. Mutation analysis was carried out using the minisequencing technique. RESULTS: The strategy was clinically applied for HLA matching in 23 cycles (20 for -thalassemia, 1 for Wiscott-Aldrich syndrome and 2 for leukaemia) from 19 couples overall involving the testing of 398 blastomeres in combination with a genetic disease, and of 6 blastomeres for HLA matching only. In 377/404 (93.3%) of blastomeres amplification products for at least one locus were produced. A total of 357 blastomeres (94.7%) gave positive amplification for all the loci investigated. Amplification failure for all the markers resulted in 27 (6.7%) blastomeres. A reliable HLA genotype was obtained in 356/377 (94.4%) of the blastomeres with positive PCR results. In total, 32 (15.8%) embryos revealed an HLA match with the affected siblings, 24 (11.8%) of them resulted unaffected and 20 (9.9%) have been transferred back to patients. In 8 (3.9%) embryos, a recombination event occurred; in 3 of these embryos the genotype was still unaffected and HLA compatible, but was not considered for transfer because of recombination event. Aneuploidy of chromosome 6 has also been observed in 17 (4.5%) of the embryos. Five clinical pregnancies were obtained, 2 of them have been already born. Two pregnancies are still ongoing and were confirmed as healthy and HLA identical with the affected children by prenatal diagnosis. One pregnancy of a b-thalassaemia case resulted in early abortion. CONCLUSIONS: Multiplex PCR of HLA STR markers has revealed a useful diagnostic tool for indirect HLA matching evaluation. The major advantage of this approach is that the validation of a single assay can be performed once and then used for the majority of the patients, reducing notably the time needed for preclinical set-up of each case. The current data confirm the feasibility of preimplantation HLA matching as a part of PGD, providing a realistic option for couples desiring an HLA-compatible child for the treatment of affected siblings.