Our group had previously established that nanoscale three-dimensional refractive index (RI) fluctuations of a linear, dielectric, label-free medium can be sensed in the far field through spectroscopic microscopy, regardless of the diffraction limit of optical microscopy. Adopting this technique, Partial Wave Spectroscopic (PWS) Microscopy was able to sense nanoarchitectural alterations in early-stage cancers. With the success of PWS on detecting cancer from healthy clinical samples, we further investigated whether and how histological staining can enhance the performance of PWS by both finite difference time domain (FDTD) simulations and experiments. In this investigation, the dispersion models of hematoxylin and eosin were extracted from the absorption spectra of H and E stained cells. Using these models, the effect of staining were studied via FDTD simulations of unstained versus stained samples with various nanostructures. We observed that, the spectral variance was increased and the spectral variance difference between two samples with distinct nanostructures was enhanced in stained samples by over 200%. Furthermore, we investigated with FDTD whether molecule-specific staining can be used to enhance signals from a medium composing of the desired molecule. Samples with two mixed random media were created and the desired medium was either stained or unstained. Our results showed that the difference between the nanostructures of only the desired medium was enhanced in stained samples. We concluded that, with molecule-specific staining, PWS can selectively target the nanoarchitecture of a desired molecule. Lastly, these results were validated by experiments using human buccal cells from healthy or lung cancer patients. This study has significant impact in improving the performance of PWS on quantifying nanoarchitectural alterations during cancer.