当前位置： SCI文献检索 > BMC CANCER期刊下所有文献 > Fractal analysis of nuclear histology integrates tumor and stromal features into a single prognostic factor of the oral cancer microenvironment.

Fractal analysis of nuclear histology integrates tumor and stromal features into a single prognostic factor of the oral cancer microenvironment.

Abstract：

BACKGROUND:The lack of prognostic biomarkers in oral squamous cell carcinoma (OSCC) has hampered treatment decision making and survival in OSCC remains poor. Histopathological features are used for prognostication in OSCC and, although useful for predicting risk, manual assessment of histopathology is subjective and labour intensive. In this study, we propose a method that integrates multiple histopathological features of the tumor microenvironment into a single, digital pathology-based biomarker using nuclear fractal dimension (nFD) analysis. METHODS:One hundred and seven consecutive OSCC patients diagnosed between 1998 and 2006 in Calgary, Canada were included in the study. nFD scores were generated from DAPI-stained images of tissue microarray (TMA) cores. Ki67 protein expression was measured in the tumor using fluorescence immunohistochemistry (IHC) and automated quantitative analysis (AQUA®). Lymphocytic infiltration (LI) was measured in the stroma from haematoxylin-eosin (H&E)-stained TMA slides by a pathologist. RESULTS:Twenty-five (23.4%) and 82 (76.6%) patients were classified as high and low nFD, respectively. nFD was significantly associated with pathological tumor-stage (pT-stage; P = 0.01) and radiation treatment (RT; P = 0.01). High nFD of the total tumor microenvironment (stroma plus tumor) was significantly associated with improved disease-specific survival (DSS; P = 0.002). No association with DSS was observed when nFD of either the tumor or the stroma was measured separately. pT-stage (P = 0.01), pathological node status (pN-status; P = 0.02) and RT (P = 0.03) were also significantly associated with DSS. In multivariate analysis, nFD remained significantly associated with DSS [HR 0.12 (95% CI 0.02-0.89, P = 0.04)] in a model adjusted for pT-stage, pN-status and RT. We also found that high nFD was significantly associated with high tumor proliferation (P < 0.0001) and high LI (P < 0.0001), factors that we and others have shown to be associated with improved survival in OSCC. CONCLUSIONS:We provide evidence that nFD analysis integrates known prognostic factors from the tumor microenvironment, such as proliferation and immune infiltration, into a single digital pathology-based biomarker. Prospective validation of our results could establish nFD as a valuable tool for clinical decision making in OSCC.

journal_name

BMC Cancer

journal_title

BMC cancer

authors

Bose P,Brockton NT,Guggisberg K,Nakoneshny SC,Kornaga E,Klimowicz AC,Tambasco M,Dort JC

doi

10.1186/s12885-015-1380-0

subject

Has Abstract

pub_date

2015-05-15 00:00:00

pages

409

issn

1471-2407

pii

10.1186/s12885-015-1380-0

journal_volume

pub_type

杂志文章

在线工具

Fractal analysis of nuclear histology integrates tumor and stromal features into a single prognostic factor of the oral cancer microenvironment.