The most simple explanation is as follows:
1.) Very low frequencies are blocked by C2 and, therefore, cannot appear at the opampoutput;
2.) Very high frequencies are fed back to the inverting opamp input with a factor of nearly "1" (via C1 and C2) - thus, they are amplified by nearly "zero".
3.) As a consequence: the circuit has a bandpass property.
4.) A somewhat more "theoretical" explanation is the folowing:
The feedback network is, in principle; a "bridget-T" network which is known to exhibit a complex "zero". Due to the feedback action of the opamp the complex zero is transferred to a complex pole pair.