It's stored in the .asc file that contains your circuit. This makes sense because you don't want to store the name for a specific part in the part file, rather anything you rename a part(C1, C2, C3, Etc) gets stored in the .asc file which contains the components and wiring of your circuit.
Edit: I understand now what you were asking. The line SYMATTR Prefix X is the line that the U is stored under. I don't know why the character X shows up as a U for the opamp but if you change that character to an A and open the model you'll see Annn instead. Every other character, including multiple characters, I changed the value to was mirrored in ltspice when i opend the .asy file.
One of your .ic spice commands probably isn't any good. That's one thing I noticed. Let me show you how to make it work. In the following image I tried to copy your schematic, add in my version of LTspice so that this detail is perfectly clear, and then show you the display I get for the current in \$L_1\$:
But the main thing I wanted to point out is that I don't think this syntax:
.ic V(vC1)=0
works in LTspice. It doesn't complain, either. So if I made that:
.ic V(QQQ)=0
LTspice also wouldn't even give me so much as a small note about it. It would just carry on and ignore it.
You don't actually have a node with that name. So LTspice just ignores it. What I did in the above schematic is to name a node so that it would work okay, giving that node the name you were trying to make work.
And it worked.
That said, I also commented out those .ic spice commands. You can see that they are blue in the schematic, now, which shows that they are comments and no longer spice commands. So LTspice is ignoring them, regardless. What I did, instead, is to add the UIC code to the end of the .tran command. You can see that in the image. This does the same thing, in effect. When you check off the "skip initial conditions" box, LTspice uses defaults for the initial conditions. Which just happen to be those values you were trying to use. So that's yet another way to do it, in this case of yours.
Bottom line is that you didn't have a node name set up correctly. So one of your initial conditions wasn't processed. That might have behaved differently in different versions of LTspice. I can't actually say, because I don't have the two version you mentioned and cannot try it out to prove my thoughts here. But you can try it out on your end and see. Label your node, as shown, and then see if both versions give you the same results. Hopefully, they do.
Best Answer
If you can format the matlab data as a series of timestamps and levels you can use the PWL file option for a voltage or current source.
http://www.linear.com/solutions/1814