It's not really possible to tell the dielectric of the SMT part without more information. It could be polyester, PET, PPS or PEN as well as possibly PP. Leakage is very low for any of those types, as is voltage coefficient, dielectric absorption (DA) is the main difference, as well as physical (temperature tolerance temperature coefficient) and, of course, availability issues in the required values and voltage ratings.
PET and PPN are more resistant to heat so more suitable for SMT capacitors, especially in lead-free processes.
A through-hole PP (assuming it's actually polypropylene and not polyester marked as PP) capacitor should be an equal or superior replacement in any case.
The very best film capacitors from a DA perspective are polystyrene, however PS softens at below 100°C so it barely suitable for careful hand soldering and completely impractical for SMT processes (maybe you could solder a socket in). A 2.2uF PS cap would also be huge and quite expensive.
If that's a coupling capacitor then DA is not highly critical but I suppose if an audio golden ear was faced with such a situation they might report sensing higher harmonics of very low frequencies where the coupling is beginning to attenuate.
When we did a lot of work with dual slope converters, the DA was quite visible on the display as rollover error, so polycarbonate (now very difficult to find) or PP (as a poor second) were used for the integrating capacitor.
If you do a web search, you can find some DIY methods of measuring DA, but if it's actually important to you I imagine you'll have a distortion analyzer or equivalent test equipment to measure the effects.