Your problem lies with the software used to create the PDF; output in the form of textboxes in a PDF is a characteristic of certain low-end PDF-creation software. There is nothing Word can do about that during the import process; you would need to clean it up afterwards.
A Word macro you could use for the clean-up is:
Sub EraseTextBoxes() Dim RngDoc As Range, RngShp As Range, i As Long With ActiveDocument For i = .Shapes.Count To 1 Step -1 With .Shapes(i) If .Type = msoTextBox Then Set RngShp = .TextFrame.TextRange RngShp.End = RngShp.End - 1 Set RngDoc = .Anchor RngDoc.Collapse wdCollapseEnd RngDoc.FormattedText = RngShp.FormattedText .Delete End If End With Next End With End Sub
Do note that whether the macro positions the output correctly depends on where the textboxes are anchored; if the anchor positions are unrelated to the textbox locations, you'll end up with a dog's breakfast. You'll probably still also end up with each line as its own paragraph. To clean up such content, see http://www.msofficeforums.com/word/29880-cleaning-up-text-pasted-websites-e-mails.html