Filtering XML documents by attributes using XSLT, retrieving elements

As a newcomer to the world of XSLT, I want to filter an XML document based on the attribute value, and then retrieve the elements containing such attributes and discard all others.

Sample XML doc

<Items>
<item1>
    <first>Lorem ipsum</first>
    <second att="abc"> dolor sit amet</title>
    <third>consectetur adipiscing elit, sed do eiusmod tempor</third>
    <forth>ncididunt ut labore et dolore</forth>
</item1>
<item2>
    <first att="def">Sed ut perspiciatis</first>
    <second>unde omnis iste natus.</title>
    <third att="ghi">error sit voluptatem accusantium</third>
    <forth>doloremque laudantium</forth>
</item2>
<item3>
    <first att="mno">At vero eos et </first>
    <second>accusamus et iust</title>
    <third>odio dignissimos ducimus qui blanditiis p</third>
    <forth att="jkl">voluptatum deleniti atque</forth>
</item3>
<item4>
    <first>Et harum quidem</first>
    <second att="abc">rerum </title>
    <third att="xyz"> soluta nobis est eligendi </third>
    <forth> Temporibus aut</forth>
</item4>

And I want to have this result (att="abc" is desired ):

<Items>
<item1>
    <first>Lorem ipsum</first>
    <second att="abc"> dolor sit amet</title>
    <third>consectetur adipiscing elit, sed do eiusmod tempor</third>
    <forth>ncididunt ut labore et dolore</forth>
</item1>

<item4>
    <first>Et harum quidem</first>
    <second att="abc">rerum </title>
    <third att="xyz"> soluta nobis est eligendi </third>
    <forth> Temporibus aut</forth>
</item4>
</items>

I've found a similar solution here but it doesn't work for me. Also, I think I should use for-each expression but I don't know how to do that.