My point has been that they clearly have an active drug, but it has schizophrenic behavior that they don’t understand. Sometimes great efficacy, zero benefit on side effects, sometimes the reverse. Sometimes a mix. In such a situation you need to know which is which w ph2 trials. Then duplicate in ph3. They violated that in CRC. They’re violating that in the TNBC ph3 since they are going first line only in one arm (ph2 was both, with first line meaningfully worse efficacy), and moon shot on second arm (effectively a completely separate trial - with its own placebo group).