Dataset Information

Tumour control probability in cancer stem cells hypothesis.

ABSTRACT: The tumour control probability (TCP) is a formalism derived to compare various treatment regimens of radiation therapy, defined as the probability that given a prescribed dose of radiation, a tumour has been eradicated or controlled. In the traditional view of cancer, all cells share the ability to divide without limit and thus have the potential to generate a malignant tumour. However, an emerging notion is that only a sub-population of cells, the so-called cancer stem cells (CSCs), are responsible for the initiation and maintenance of the tumour. A key implication of the CSC hypothesis is that these cells must be eradicated to achieve cures, thus we define TCPS as the probability of eradicating CSCs for a given dose of radiation. A cell surface protein expression profile, such as CD44high/CD24low for breast cancer or CD133 for glioma, is often used as a biomarker to monitor CSCs enrichment. However, it is increasingly recognized that not all cells bearing this expression profile are necessarily CSCs, and in particular early generations of progenitor cells may share the same phenotype. Thus, due to the lack of a perfect biomarker for CSCs, we also define a novel measurable TCPCD+, that is the probability of eliminating or controlling biomarker positive cells. Based on these definitions, we use stochastic methods and numerical simulations parameterized for the case of gliomas, to compare the theoretical TCPS and the measurable TCPCD+. We also use the measurable TCP to compare the effect of various radiation protocols.

SUBMITTER: Dhawan A

PROVIDER: S-EPMC4014481 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Tumour control probability in cancer stem cells hypothesis.

Dhawan Andrew A Kohandel Mohammad M Hill Richard R Sivaloganathan Sivabal S

PloS one 20140508 5

The tumour control probability (TCP) is a formalism derived to compare various treatment regimens of radiation therapy, defined as the probability that given a prescribed dose of radiation, a tumour has been eradicated or controlled. In the traditional view of cancer, all cells share the ability to divide without limit and thus have the potential to generate a malignant tumour. However, an emerging notion is that only a sub-population of cells, the so-called cancer stem cells (CSCs), are respons ...[more]

PMID: 24811314

Similar Datasets

Project description:PurposeTo evaluate the applicability and estimate the radiobiological parameters of linear-quadratic Poisson tumour control probability (TCP) model for primary prostate cancer patients for two relevant target structures (prostate gland and GTV). The TCP describes the dose-response of prostate after definitive radiotherapy (RT). Also, to analyse and identify possible significant correlations between clinical and treatment factors such as planned dose to prostate gland, dose to GTV, volume of prostate and mpMRI-GTV based on multivariate logistic regression model.MethodsThe study included 129 intermediate and high-risk prostate cancer patients (cN0 and cM0), who were treated with image-guided intensity modulated radiotherapy (IMRT)?±?androgen deprivation therapy with a median follow-up period of 81.4 months (range 42.0-149.0) months. Tumour control was defined as biochemical relapse free survival according to the Phoenix definition (BRFS). MpMRI-GTV was delineated retrospectively based on a pre-treatment multi-parametric MR imaging (mpMRI), which was co-registered to the planning CT. The clinical treatment planning procedure was based on prostate gland, delineated on CT imaging modality. Furthermore, we also fitted the clinical data to TCP model for the two considered targets for the 5-year follow-up after radiation treatment, where our cohort was composed of a total number of 108 patients, of which 19 were biochemical relapse (BR) patients.ResultsFor the median follow-up period of 81.4 months (range 42.0-149.0) months, our results indicated an appropriate ?/??=?1.3 Gy for prostate gland and ?/??=?2.9 Gy for mpMRI-GTV. Only for prostate gland, EQD2 and gEUD2Gy were significantly lower in the biochemical relapse (BR) group compared to the biochemical control (BC) group. Fitting results to the linear-quadratic Poisson TCP model for prostate gland and ?/??=?1.3 Gy were D50?=?66.8 Gy with 95% CI [64.6 Gy, 69.0 Gy], and ??=?3.8 with 95% CI [2.6, 5.2]. For mpMRI-GTV and ?/??=?2.9 Gy, D50 was 68.1 Gy with 95% CI [66.1 Gy, 70.0 Gy], and ??=?4.5 with 95% CI [3.0, 6.1]. Finally, for the 5-year follow-up after the radiation treatment, our results for the prostate gland were: D50?=?64.6 Gy [61.6 Gy, 67.4 Gy], ??=?3.1 [2.0, 4.4], ?/??=?2.2 Gy (95% CI was undefined). For the mpMRI-GTV, the optimizer was unable to deliver any reasonable results for the expected clinical D50 and ?/?. The results for the mpMRI-GTV were D50?=?50.1 Gy [44.6 Gy, 56.0 Gy], ??=?0.8 [0.5, 1.2], ?/??=?0.0 Gy (95% CI was undefined). For a follow-up time of 5 years and a fixed ?/??=?1.6 Gy, the TCP fitting results for prostate gland were D50?=?63.9 Gy [60.8 Gy, 67.0 Gy], ??=?2.9 [1.9, 4.1], and for mpMRI-GTV D50?=?56.3 Gy [51.6 Gy, 61.1 Gy], ??=?1.3 [0.8, 1.9].ConclusionThe linear-quadratic Poisson TCP model was better fit when the prostate gland was considered as responsible target than with mpMRI-GTV. This is compatible with the results of the comparison of the dose distributions among BR and BC groups and with the results achieved with the multivariate logistic model regarding gEUD2Gy. Probably limitations of mpMRI in defining the GTV explain these results. Another explanation could be the relatively homogeneous dose prescription and the relatively low number of recurrences. The failure to identify any benefit for considering mpMRI-GTV as the target responsible for the clinical response is confirmed when considering a fixed ?/??=?1.6 Gy, a fixed follow-up time for biochemical response at 5 years or Gleason score differentiation.

Project description:The biological complexity reflected in histology images requires advanced approaches for unbiased prognostication. Machine learning and particularly deep learning methods are increasingly applied in the field of digital pathology. In this study, we propose new ways to predict risk for cancer-specific death from digital images of immunohistochemically (IHC) stained tissue microarrays (TMAs). Specifically, we evaluated a cohort of 248 gastric cancer patients using convolutional neural networks (CNNs) in an end-to-end weakly supervised scheme independent of subjective pathologist input. To account for the time-to-event characteristic of the outcome data, we developed new survival models to guide the network training. In addition to the standard H&E staining, we investigated the prognostic value of a panel of immune cell markers (CD8, CD20, CD68) and a proliferation marker (Ki67). Our CNN-derived risk scores provided additional prognostic value when compared to the gold standard prognostic tool TNM stage. The CNN-derived risk scores were also shown to be superior when systematically compared to cell density measurements or a CNN score derived from binary 5-year survival classification, which ignores time-to-event. To better understand the underlying biological mechanisms, we qualitatively investigated risk heat maps for each marker which visualised the network output. We identified patterns of biological interest that were related to low risk of cancer-specific death such as the presence of B-cell predominated clusters and Ki67 positive sub-regions and showed that the corresponding risk scores had prognostic value in multivariate Cox regression analyses (Ki67&CD20 risks: hazard ratio (HR) = 1.47, 95% confidence interval (CI) = 1.15-1.89, p = 0.002; CD20&CD68 risks: HR = 1.33, 95% CI = 1.07-1.67, p = 0.009). Our study demonstrates the potential additional value that deep learning in combination with a panel of IHC markers can bring to the field of precision oncology.

Dataset Information

Tumour control probability in cancer stem cells hypothesis.

Publications

Tumour control probability in cancer stem cells hypothesis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets