Generally, there is no difference for the LibTiff.Net if you image is a 10 bits per sample or 8, or 16.
Some TIFF flavors can not be converted to RGBA using ReadRGBAImage
method but the library should be able to decode them anyway.
Have a look at Basic introduction to the capabilities of the library article. It contains enough information about different approaches to reading images.
In you case you might want to use ReadScanline
or ReadEncodedStrip
/ ReadEncodedTile
methods. Using these you should get decoded image data and then it is up to you how you will interpret the data.
You might want to check byte ordering, too. If you don't, you might end up with color channels in wrong order.