Here's one option, to give you the idea. Note that the values are in
quotes because your Var5 is a factor. If it isn't supposed to be, you
may want is.null() in the same way. I included NA because you asked,
but there aren't any NA values in Var5.
Maybe something like the following is what you want.
The code first creates a logical index of columns with at least one NA
or "NULL" (character string, not NULL) values. Then extracts only those
columns from the dataframe.