>Why have radiolabeled mAbs not made more headway in the treatment of solid cancers?<
My guess would be due to AE considerations.
The subset of surface markers that are the target of drugs for hematological cancers are relatively selective in that you're not apt to have many "solid" organs that express them. So one could argue that Zevalin and Bexxar would not create much collateral damage to other organs (save for the ones that process / degrade them).
When you get to antibodies that target cell surface proteins on solid cancers, these proteins (EGF-R for example) are invariably expressed to some degree on other healthy organs (versus "overexpressed" on the cancer) and therefore you'd be more prone to exacerbating what are currently manageable AEs. It would be the difference between Erbitux giving the patient a rash because it saturates receptors at the skin, versus a radioactive Erbitux giving patients a rash and a deposit of radioactivity all along the surface of their body. In one case it is a predictable and manageable side effect, in the other it would seem entirely undesirable and a deal breaker. That would be my guess.
Of course this doesn't explain why an antibody like Avastin, against a floating cytokine, isn't radioactively tagged to try and improve efficacy. Perhaps the answer in that circumstance is the well documented (and disappointing) market experience with RIT, and therefore a reluctance on the part of the sponsor to spend much time and effort for little monetary gain.