The simple answer is to recognise that C1/R1 and 1Meg/13pf is a special case, given the application : it is almost certainly a 10:1 oscilloscope probe.
1 Meg and some small C is a standard scope channel input : the precise value of C is undefined but happens to be 13 pF fro this scope.
Then R1 is 9 Megohms to give the correct DC attenuation, and C1 is adjusted for a flat frequency response (actually, an accurate square wave response) each time you plug that particular 10:1 probe into that particular scope. (Calibrating the probes is just a routine part of using a scope. The scope will provide a "CAL" output for that purpose, and the correct plastic trimming tool lives in that pouch on top of the scope.)
The probe output is then a perfectly scaled version of the probe input, with the same frequency response. Which means you can model the entire circled block as Cp in parallel with 1.3 pF and 10 megohms (followed by a perfect 20dB attenuator)
Given the source impedance of 50R, you can for all practical purposes ignore the 10 megohm probe resistance, giving an attenuator composed of 50R and (Cp + 1.3 pF) to which you now have to add the impedance of the cable.