Update and complete solution for pdf rendering with Special characters
If the PDF file fails to display all the page’s text, particularly multibyte characters such as Japanese or accented international characters, adjust your CSS to use a font that supports them.
Here escape
attribute doesn't matter. only following style for font matters where that particular character code supports.
<style>
body { font-family: 'Arial Unicode MS'; }
</style>
Visualforce
<apex:page standardController="Opportunity" renderAs="pdf" applybodyTag="false"
showHeader="false">
<head>
<style>
body { font-family: 'Arial Unicode MS'; }
</style>
<div>
{!myvalue}
</div>
<apex:outputText value="{!myvalue}" escape="false" />
</apex:page>
Results:

Refer Visualforce PDF Rendering Considerations and Limitations
{!myString}
syntax all work for me. Are you sure that you've used UTF-8 encoding when developing your page/controller? If not, that could be your problem.