The concept of Enterprise Application Integration (EAI) is widely used for integrating heterogeneous applications and systems by message-based communication. Typically, EAI servers provide a huge set of specific inbound and outbound adapters used for interacting with the external systems and for converting proprietary message formats. However, the main problems in currently available products are the monolithic design of these adapters and performance deficits caused by the need for data independence. First, we classify and discuss these open problems. Second, we introduce our model-driven DIEFOS (data independence, efficiency and functional flexibility using feature-oriented software engineering) approach and show how the feature-based generation of dynamic adapters can improve data independence, efficiency and functional flexibility. Finally, we analyze open research challenges we see in this context.