Nanozoomer digital pathology scanner

Frozen liver samples were embedded in OCT and cut in 10 μm sections. Sections were brought to room temperature, fixed with 10% buffer formalin for 5 min, washed with 60% isopropanol, then saturated with Oil Red O (1% w/v, 60% isopropanol) for 15 min, washed in 60% isopropanol and rinsed in distilled water. Then sections were then mounted in aqueous mounting medium with coverslips. Images were acquired using a NanoZoomer Digital Pathology Scanner (Hamamatsu Photonics K.K., Japan) and analysed using the NDP Viewer software.

Assessment of immunohistochemical staining was conducted at 200x magnification following high resolution scanning using a Nanozoomer Digital Pathology Scanner (Hamamatsu Photonics). Staining in the cytoplasm was assessed using a semi-quantitative immunohistochemical H score; where staining intensity was assessed as none (0), weak (1), medium (2) or strong (3) over the percentage area of each staining intensity. Nuclear staining was assessed as the percentage of nuclei with any percentage intensity of staining. Greater than 30% of cores were double assessed, with both assessors blinded to clinical outcome and each other's scores. The single measure intraclass correlation coefficient were above 0.7, indicating good concordance between scorers.
Statistical analysis was performed using IBM SPSS Statistics (version 24). Data was stratified based on breast cancer specific survival using X-Tile software [33 (link)]. All differences were deemed statistically significant at the level of P<0.05. The Pearson χ² test of association was used to determine the relationship between categorised protein expression and clinicopathological variables. Survival curves were plotted according to the Kaplan-Meier method with significance determined using the log-rank test. The primary endpoint of this study was to determine if GPER expression is associated with breast cancer specific survival.

Specimens were decalcified in 10% ethylenediaminetetraacetic acid for 3 weeks at room temperature while being subjected to agitation on a shaker. Specimens were dehydrated in a series of ethanol solutions with concentrations ranging from 75% to 100%, embedded in a paraffin-embedding machine (TKD-BMC, Hubei, China), and sliced into 5-µm-thick posterior sections using a microtome (RM2235, Leica, Germany). Following that, hematoxylin-eosin (HE) staining, Masson’s trichrome staining, saffron-solid green (S&G) staining, and tartrate-resistant acid phosphatase (TRAP) staining were performed separately according to the manufacturer’s instructions. The images were scanned using a NanoZoomer digital pathology scanner (NDP; Hamamatsu, Japan). Three slices were randomly selected based on HE staining, and new bone areas were automatically measured and calculated using Image-Pro Plus 6.0 (IPP). The formula used to calculate new bone area is as follows: New bone area (%) = new bone area/tissue area × 100%.

The brightfield images of the chromogenic ALK IHC stainings were digitalized on a NanoZoomer Digital Pathology scanner (Hamamatsu, Japan). For IF we used a Zeiss AxioImager M2 m microscope and an Olympus slide scanner VS120-L100. Each image was taken in both DAPI (Alexa350) and Cy5 (Alexa647) channels. For all images within a figure, the same parameters of acquisition, such as filter set, exposure time, and filter intensities were used.

Staining was assessed at × 200 magnification following high-resolution scanning (Nanozoomer Digital Pathology Scanner, Hamamatsu Photonics). Assessment was conducted with involvement and training provided by a consultant neuropathologist (SP). Cytoplasmic staining was semi-quantitatively assessed using an immunohistochemical H-score [27 (link)], where staining intensity was assessed as none (0), weak (1), medium (2), or strong (3) over the percentage area of each staining intensity. Nuclear staining was assessed as the percentage of nuclei with any intensity of staining. 30% of cores for each protein were examined by a second independent assessor blinded to clinical outcome and the primary assessor’s scores. Good concordance was demonstrated between scorers (single measure intraclass correlation coefficients were above 0.7 for all markers assessed across all brain tumour cohorts) (Supplementary Table 4). Unbiased cut-points, for stratification, were obtained based on overall survival using X-tile software [28 (link)].