Indeed. We see these kind of n=1 reports every now and then and media jump on the hype gun without understanding drug development at all.

Statistically and probabilistically speaking, we can make an estimate (with error bars) of the value of the treatment, even with a trial of only a few people.

it was a trial of two, and the success ratio was 50%. so there is some actual statistical data there.

Take a look at http://www.mayoclinicproceedings.org/article/S0025-6196(14)0... and the included video.

The first patient did have a recurrence of a single tumor that was treated locally, with her marrow remaining completely clear. The second patient had his marrow cleared and all but a couple tumors in his legs resolved. And the virus penetrated those tumors and allowed full 3D mapping of them due to tags attached to the virus itself (it sounded like a thyroid produced chemical was attached to the virus as a tag). The implication was that because of the detailed mapping that it would be possible to resolve/remove the remaining tumors in the leg muscles of the second patient.

So, the answer is quite a bit much more complex than 50%. Both patients cleared the myeloma from their marrow. Almost all (numerous!) tumors were removed, from a single treatment in each patient.

Edit: Here's a link to the editorial article that accompanies the paper: http://www.mayoclinicproceedings.org/article/S0025-6196(14)0...

yeah so in that case 'a trial of 1 proves nothing' is especially misleading. this seems like very strong evidence.

What happens if it turns out that the treatment is fatal in 30% of cases? A trial of 2 would not necessarily show this type of thing. That is why one cannot extrapolate this without much more information i.e. more trials.

nobody is saying that more trials are unnecessary. my point is to emphasize the limited data so far is compelling, and that it goes beyond a 'trial of 1'.