The Laser Interferometer Space Antenna (LISA) is a space mission, planned to be launched in 2034, to observe gravitational waves in the highly promising frequency band between 100 μHz and 0.1 Hz. To demonstrate its technical feasibility, the LISA Pathfinder (LPF) space mission was operated from December 2015 until July 2017. LPF successfully showed that a Test Mass (TM) can be set into free fall and that its displacement to another suspended TM can be measured with laser interferometers, both sensitive enough to be applicable for LISA. Further questions and technical challenges, however, need to be addressed on the path towards LISA.One difference between the two missions is their duration. The much longer LISA mission will have to cope with a higher dose of cosmic radiation. Hence, the long-term stability of radiation sensitive components is an important parameter under test. For LPF, the photoreceivers, InGaAs Photodiodes (PDs), were expected to significantly decrease in responsivity by 17%. As a consequence, the stabilities of the actual in-flight PD responsivities were monitored, to compare with the pre-flight estimates, as described in this thesis. The responsivity measurement experiment used an independent optical power measurement to calibrate the PDs. The signal was generated by a power modulation of the laser beam, inducing a radiation pressure modulation at the TMs. The corresponding differential TM displacement was measured with the precise interferometric readout, thus giving a calibration of the laser power. It was found that the PDs on LPF did not degrade within the statistical errors of 1% during the full monitoring duration of one year. Therefore, in a simplified linear extrapolation to a 6 year LISA mission, a decrease in the antenna sensitivity of more than 6%, as consequence of less detected power, is not expected.The analysis of individual beam powers revealed an unexpected low frequency power noise that became the second main research question of this thesis. With a dedicated model of the optical interferometer paths, the noise source could be identified as an unstable polarisation state with up to 4.5% off-nominal polarised power. Nominally, the polarisation on LPF was cleaned with a Polarising Beam Splitter (PBS). However, a laboratory analysis of flight spare PBSs showed a decrease in the polarisation purity, which originated from slow out-gassing ofwater when exposed to vacuum for three weeks. An extrapolation of the observed effect for a longer exposure to vacuum, in combination with a degraded incident polarisation state to the PBS were found to be the best explanation for the in-flight polarisation instabilities.The impact of the unexpected noise on LPF’s scientific results was found to be negligible, with the exception of the induced radiation pressure to the TMs below 0.1mHz. For the longest noise run in February 2017, the best estimate of the radiation pressure contribution to the differential TM acceleration noise amplitude was found to be up to 26 (+6 - 2)% at 74 μHz.As a consequence for LISA, the sensitivity to polarisation should be considered during the selection process of optical components and their long-term performance in a space environment should be tested.