Adriana Mosch 

When Trados opens a PDF it actually does convert it to a DOCX in the background, and the file you are working on is a DOCX.  I'd recommend you take the DOCX, which will be in the same folder as the source file now, and then edit it to tidy it up.  Then translate the DOCX.

Adriana Mosch 

The advice given by Paul is OK, but will need a lot of work. Better solution for ANY PDF translation is always to convert the PDF file first to Word using either Acrobat DC full version (works for many PDFs very good) or a decent OCR solution, like Abbyy FineReader. The conversion must be done manually to avoid problems you encountered. Only manually you can see, if the recognition area has been properly assigned, for example if two-column text is treated as such or as table and not like in your case simply separated by tabs or even worse by spaces.

Paul Jerzy Czopik  Thank you for your solutions. I'll try them.

So, I tried it the way Paul suggested and edited the .docx file to tidy it up. It worked out quite well. Thank you.