The ice immediately around the islands looks a little thinner than the body of the pack. Looking at ice motion maps, you can see why this happens, for example this one from the US Navy. That's an animation of the ice thickness/motion for the last 30 days: the New Siberian Islands are at ~140E, ~75N.
As the pack moves around, it leaves open water on the leeward side of the islands, which immediately freezes over. On the opposite side of the island, the ice gets piled up much thicker. As the direction of ice movement changes, you end up with a ring of thinner ice around each island, then a ring of thick ridged ice, then finally the main body of the Arctic pack ice.
That opening in the ice isn't a true polynya, and was a very transient feature. During the Arctic re-freeze, two things occur: outwards expansion of the central pack, and formation of shore ice along the coastline. As these two areas of ice approach each other, you get what looks like a polynya, but which is really simply part of the area exposed by the summer melt that hasn't yet refrozen.
Compare these photos:
October 19-25: the central pack expands and the shore ice starts freezing up.
October 25-31: a small area gets pinched between the two...
October 31-Nov 3: ... and promptly freezes over a couple of days later.
It wasn't really a polynya, and it certainly wasn't persistent. It was an area of delayed re-freeze, brought about by an early thaw and consequent extra absorption of energy in the top layers of water.
Please see image on page 9 at this link: http://www.flipdocs.com/showbook.aspx?ID=10004692_698290
Recent MODIS Rapidfire image of the East Siberian Arctic Shelf (I am not going to vouch for these things -- I really have no idea what I am looking at):