How can I merge rows from the same dataframe with a mix of null values

Question

This problem started when I used pivot_table on my initial table that contained:

col_1         col_2      col_3  col_4   col_5   year    index     value
Company X