I think the HCV replicon assay does have some serious limitations , but it also has a decent record of predicting in vivo results so it's been widely used to screen preclinical candidates. Newer models , including an infectious cell-culture system where the cells actually produce complete virions that can infect new cells ( or chimps ) , and one using immune-deficient mice with humanized livers , will probably overcome some of the limitations of the replicon assay. Still , they're just models. If the 283/riba interaction paper had been about studies in chimps , it would probably be 'game over' , but since the replicon assay was used we still have some reason to be hopeful.
Also , as we've discussed before , there may be workarounds to a 283-riba interaction. If a triple combo of 283/PI/ifn had potent antiviral activity , e.g. achieving HCV-negativity in a week or so and maintaining it for 10-12 wks. , you might be able to follow that with a triple of PI/ifn/riba , or just ifn/riba , to consolidate the immune response and allow SVR. It wouldn't really be too much different from what VRTX will be doing with their 12 plus 12 approach. It all boils down to whether riba is critical in that first stage of treatment.
A negative interaction study would likely put IDIX in a weak negotiating position on 283/PI combo approaches , since IDIX would need the PI more than the other company would need 283.
OTOH , IDIX could surprise us with that " creative approach " they've been hiding from us. ;)