>> You imply that you can extrapolate to 4 SVR in 4x34 pts <<
Actually , I'm not implying that. I'm implying that you can extrapolate to something >4.
You don't need to know statistics to opine on this , just what dozens of readily available abstracts will tell you about expected SVR rates in such a nonresponder population , which includes "relapsers" as opposed to strictly null responders , who are treated with up to 72 weeks of pegifn / riba.
Make a weekend project out of it and let us know what you find out.