how to read all columns but the one use for partition

[sdementen]

I am storing timeseries dataframes (index=datatimeindex, multiple columns of data).
I add a column "year" with the df.index.year.
I write to the collection with collection.write(item_name, df, overwrite=True, partition_on=["year"]).
When I read it back, I use item = collection.item(item_name, filters=[("year", "==", year)]) and I would like to avoid reading (for performance) the "year" column (as it is only used for partitioning). I can read the columns in item.data.columns and remove from this Index the "year". But then, in the item.to_pandas(), I cannot specify the columns to read from.
Any way to do what I want to do properly ?