Does heterogeneity matter in the estimation of tumour budding and tumour stroma ratio in colon cancer?

Abstract

Background: Tumour budding (TB) and Tumour Stroma Ratio (TSR) may be rewarding in the treatment stratification of patients with stage II colon cancer. However, lack of standardization may exclude these parameters from being used in a clinical setting. The purpose of this methodologic study was to compare stereology with semi-quantitative estimations of TSR, to investigate the intra-tumoural heterogeneity of TB and TSR, and to assess the intra- and inter-observer agreement.

Methods: Three paraffin embedded tumour blocks, one of them representing the deepest invasive front, were selected from each of 43 patients treated for stage II colon cancer. TSR was estimated in H&E sections semi-quantitatively using conventional microscopy, and stereologically on scanned slides, using the newCAST stereology platform. TB was scored across 10 high power fields at the invasive front in cytokeratin AE1/AE3 stained sections.

Results: Subjective, semi-quantitative estimates of TSR significantly correlated to the stereological estimates, with the best correlation found for sections with the deepest invasive tumour penetration (σ = 0.621, p < 0.001). Inter-observer agreement was moderate to substantial for both TB (Κappa = 0.46-0.73) and TSR (Κappa = 0.70-0.75). The Intraclass correlation coefficient (ICC) for TSR varied from 0.322 based on stereological hotspot estimation to 0.648 for the semi-quantitative evaluation. For TB, ICC varied from 0.646 based on continuous data to 0.698 based on categorical data (cut-off: 10 buds). Thus, the intra-tumoural heterogeneity for both TB and the semi-quantitative estimation of TSR was low.

Conclusion: We recommend using only one tissue section representing the deepest invasive tumour area for estimation of TSR. For TB we recommend using one tissue section; however due to low representation of high-budding tumours, results must be considered with caution.

Keywords: Colon cancer; Heterogeneity; Tumour budding; Tumour stroma ratio.

Fig. 1

Stereological estimation of tumour stroma ratio. a The area with the highest stroma density (yellow frame) is subjectively selected at low magnification (2×). b At higher magnification (10×) a grid of 4 x (6 × 7) points is superimposed on the selected area. Points hitting either tumour cells or stroma are counted separately. Tumour cells must to be present at all borders of the visual field (north-east-south-west)

Fig. 2

Counting tumour budding in a high power field (HPF). The number of tumour buds (TB) was counted along the invasive front on cytokeratin AE1/AE3 stained sections in 10 HPF. A tumour bud was defined as an isolated single tumour cell or a cluster of up to four tumour cells (arrows). Adenocarcinoma cells without a clear nucleus and cytoplasmatic fragments (arrowheads) were not counted. Magnification 40×

Fig. 3

How to distinguish fibrous tumour stroma from smooth muscle tissue. a In the H&E stained section it is challenging to distinguish between fibrous stroma and smooth muscle tissue. b Immunohistochemical stain for desmin highlights the smooth muscle tissue. c Masson trichrome highlights the collagen fibres in the fibrous stroma (blue) and the smooth muscle tissue (purple-red). Magnification 10×