What can we learn from the breaking of the Wandzura-Wilczek relation?

We review the study the Wandzura-Wilczek relation for the structure function g_2, with a particular attention on the connection with the framework of Transverse Momentum Dependent factorization. We emphasize that the relation is broken by two distinct twist-3 terms. In the light of these findings, we clarify what can be deduced from the available experimental data on g_2, which indicate a breaking of the order 20-40%, and how to individually measure the twist-3 terms.