Is there any evidence for immunologically mediated or immunologically modifiable early pregnancy failure?

Purpose: Human reproduction is an inefficient process. There is a high rate of loss of early pregnancies, often before the mother (or physician) knows she is pregnant. Genetic abnormalities can explain much of the wastage, but can it explain all of the failures? As embryos bear paternal and embryonic antigens foreign to the maternal immune system, could some otherwise normal embryos be "rejected"? Methods: Critical review of existing data. Results and Conclusions: Otherwise normal embryos can fail prior to implantation, at implantation, in the periimplantation period as occult/chemical pregnancies, and as clinically evident miscarriages. The maternal immune system and its products (e. g., cytokines) can have innocent bystander effects, and a good case for direct recognition and "rejection" can also be made. The tools needed for accurate clinical diagnosis of such situations require further development and validation. Deliberate modification of the maternal host defence system can improve the chance of success, but the best evidence for efficacy of immunotherapeutic interventions is the situation of recurrent spontaneous abortions, which constitutes only a small percentage of losses. There is also evidence of clinical efficacy for several types of treatment to improve implantation and early pregnancy success.