Sorry, but the K/M chart you just posted is impossible (no math can create it).
The 7 events at 35 months dropped the curve about 25%, implying 28 or so were at risk, yet only 2 censors (and no events) were post that point.
OK, maybe you want to add 20 more censors in at 35 months and did not show it well? No math there either, as you they did not enroll 60 patients in a single month.
Further, it should be obvious that you have a lack of censors from 20-27 months on the treatment arm. That means you think all (but 1) that enrolled in summer/fall '15 have evented. Strange, I guess the German "optimization" is killing patients.
All you are doing is trying to draw K/M charts that show what you want to show. That is not a model.