I agree that the halt had nothing to do with a safety signal. This much was certain then, and has since been confirmed by ASCO 2017 presentation of overall safety experience.
I disagree with your assertion, however, that the halt indicates a failed futility signal on PFS. In my experience, the regulators (in this case later confirmed to be FDA) do not step in to halt a trial for failure at futility unless the DSM indicates that treatment may be harming patients (I.e., performing worse than placebo). A company has many options to take and advantages to gain in continuing a trial that may ultimately fail. This also aligns BTW with the above assertion that halt had nothing to do with safety, for this would be the justification to override company judgement. I may be wrong here, but I believe it is highly, highly doubtful that PFS failure explains the halt.
My belief which has been stated here before is just the opposite which is that a statistically significant positive PFS signal was seen upon a June-July 2015 futility review and that the company refused to comply with a DSM recommendation to stop the trial prematurely. Now, management must show a positive OS to justify delaying results by over 2 years, diluting my interest 7 fold since then, and delaying development of DIRECT.