This work shows a cascaded internal multiple attenuation workflow based on top-down inverse scattering series (ISS) predictions followed by adaptive subtraction. The ISS multiple modeling is purely data driven and does not assume a priori subsurface information such as velocity field and known generating horizons. Adaptive subtraction is employed to match the predicted model with the internal multiples present in the data set. A case study using the top- down cascaded workflow was applied to a pre-stack field data set located in western Alberta, Canada, where the interval between the Duvernay formation and the shale/basin contact is contaminated by strong internal multiples. The workflow attenuated most of the internal multiples present in the target zone, improving the overall primary resolution and highlighting weak events previously hindered by multiples.