I can't give you what the title promises, not exactly. But I can get into neighborhood using what has been published already. You can find the source for sorafenib Kaplan-Meier survival curves in TIVO-1 (tivozanib vs. sorafenib) trial here . I digitized sorafenib KM curves from figures 2 (Sorafenib Plus Tivozanib 60 %), 6A (Sorafenib Plus Tivozanib 100 %) and 7A (Sorafenib Plus Tivozanib 0 %). Figure 6A is actually Sorafenib plus any anticancer therapy 100%, but tivozanib is 92.8 % of that any anticancer therapy, other targeted therapies are about 3.5 % and the rest generally ineffective treatments - tivozanib 100% is therefore an approximation that very likely puts tivozanib on disadvantage. The group Sorafenib Plus Tivozanib 100 % has 168 patients, the group Sorafenib Plus Tivozanib 0 % has 89 patients AND Sorafenib Plus Tivozanib 60 % has 257 patients. 168 + 89 = 257. We have Sorafenib from overall survival comparison (Figure 2 in ) split into two legitimate subgroups. Figure 1 below confirms that.
Slight differences between the curves in Figure 1 are due to my digitizing errors (heavy caffeine shakes compounded with those due to nicotine. Yes, I know: bad habits) and the approximate nature of KM curves. The more interesting action is in Figure 2.
Figure 2 is an almost unfair, inaccurate comparison. Why? Well, the group Sorafenib Plus Tivozanib 100% is not similar to the group Sorafenib Plus Tivozanib 60 % which in turn is not similar to the group Sorafenib Plus Tivozanib 0 % which is not similar to the group Sorafenib Plus Tivozanib 100 %. Why not? Different MSKCC risk group distributions at the starting blocks i.e. at time zero. We do know what the MSKCC risk group distribution is for the group Sorafenib Plus Tivozanib 60 % but not for the other two. Am I then wasting my and, more importantly, your time? Not really. I can state something factual on both groups with unknown MSKCC risk group distribution.
- The group Sorafenib Plus Tivozanib 100 % has MSKCC distribution that puts it in disadvantage when compared to the group Sorafenib plus Tivozanib 60 %. It percentage of MSKCC Favorable (the long surviving ones) is lower than the corresponding percentage for Sorafenib Plus Tivozanib 60 %. That should be obvious; the Favorable group survives the longest, so they do not cross over in high numbers.
- The group Sorafenib Plus Tivozanib 0 % has MSKCC risk group distribution that gives it advantage over the group Sorafenib Plus Tivozanib 60 %. This is exactly for the same reason as above. MSKCC risk group Favorable survives the longest hence they will be present more pronouncedly in the group Sorafenib Plus Tivozanib 0 %.
Adding up 1 plus 2 gives: a fair comparison would be even worse for Sorafenib without subsequent Tivozanib (my group Sorafenib Plus Tivozanib 0 %).
The flattening of the curve for the group Sorafenib Plus Tivozanib 0 % is probably caused by the number of at risk population dropping below 50 which will make survival curve inaccurate (time between deaths increase and drops due them become bigger). For comparison how a larger group might have behaved I have added the overall survival curve for sorafenib from the original sorafenib trial  (eerily close? Isn't it?). It has at risk population well over 200 around 50 % to 60 % survival. See also the presentation for axitinib vs. sorafenib and temsirolimus vs. sorafenib trials  if you want to see how overall survival curves look when the at risk population drops below 50.
A message to FDA. Following your example I did compare survival curves for Sorafenib Plus Tivozanib 0 % and Sorafenib Plus Tivozanib 60 % at 60 % survival point. Guess what: difference seems to be about 8 to 9 months favoring the latter. What do you make out about that? A statement about late and unobservable toxicity of sorafenib? There is no tivozanib in 0 % tivozanib unless it is some kind of ninja zero that isn't actually zero.
The Grand Finale:
Where you do think the overall survival curve for tivozanib from TIVO-1 trial would go in Figure 1; Pin the Tail in the Donkey, please.
- Overall survival results from a Phase III study of tivozanib hydrochloride vs sorafenib in patients with renal cell carcinoma
- Sorafenib for Treatment of Renal Cell Carcinoma: Final Efficacy and Safety Results of the Phase III Treatment Approaches in Renal Cancer Global Evaluation Trial
- 2012 European Oncology Congress in Vienna, Clinical Spotlight in Renal Cell Carcinoma, Abstract LBA22, Abstract 793PD
Disclosure: I am long AVEO.
Additional disclosure: And Still Not Selling.