The findings illuminate how prostate cancer circulating tumor cells negatively affect patient outcomes, presenter Mohammad Alhusseini, PhD, of the University of Miami Miller School of Medicine in Florida noted.

"Circulating tumor cells are cancer cells detectable in the peripheral blood released by the tumor," he said. "[They] constitute seeds for subsequent growth of additional tumor metastases, triggering a mechanism that is responsible for many cancer-related deaths."

The ISMRM is holding this week's meeting in conjunction with the European Society for Magnetic Resonance in Medicine and Biology and the International Society for MR Radiographers and Technologists.

An indicator of prostate cancer risk

The number of circulating tumor cells (CTCs) is an indicator of metastatic risk in prostate cancer and is a meaningful biomarker, the team explained.

"Quantitative multiparametric MRI features in prostate cancer are associated with tumor aggressiveness," Alhusseini said.

The investigators explored any links between radiomic features from multiparametric MRI of the prostate and circulating tumor cell counts in 73 prostate cancer patients. They developed and trained a neural network algorithm to predict circulating tumor cell counts (with a threshold of five or more CTCs indicating risk of metastatic disease). The team conducted 100 training and testing runs of five-fold cross-validation, measuring results using the area under the receiver operating curve (AUC).

The team used the following radiomics features to train the algorithm:

1. "T-onset" -- time of onset of the contrast wash-in
2. T2 intensity, 50%
3. Higher tumor volume, measured by a tool the group developed for the algorithm called the Habitat Risk Score 6 (HRS6), which is based on a one to 10 scale ("Volumes with HRS6 are concordant with tumor volumes from radical prostatectomy samples," the authors wrote in their abstract.)
4. Apparent diffusion coefficient (ADC) intensity skewness
5. ADC contrast mean
6. ADC contrast kurtosis
7. B-value intensity

Alhusseini's group collected blood from study participants enrolled in two trials: one an "active surveillance" trial and another a randomized radiotherapy trial that investigated the effects of an MRI-guided prostate radiotherapy treatment. Patients underwent multiparametric MRI (including axial T2-weighted MRI, diffusion-weighted imaging [DWI] and dynamic contrast-enhanced [DCE]-MRI exams) and MRI-ultrasound fusion biopsy. The amount of circulating tumor cells was assessed from the blood draws.

In the active surveillance trial arm, the mean circulating tumor count per 7.5 mL of blood was 19. In the radiotherapy trial arm, the median circulating tumor count in 7.5 mL of blood was 74.

Using the seven features, the team found an average AUC of 0.769 for a connection with higher CTC counts and adverse effects; of the seven, HRS6, lower T2 intensity measurements, and time of onset were the top three, with an AUC of 0.834.

"To the best of our knowledge, this is the first study to model CTC counts with in vivo multiparametric MRI radiomics," the group concluded. "Magnetic resonance imaging radiomics features are promising markers of prostate cancer metastatic risk."

BJR publishes special focus on prostate cancer
The British Journal of Radiology (BJR) has dedicated its latest issue to a diverse collection of prostate cancer articles.

U.K. researchers get cash injection for MRI prostate cancer trial
A team from Imperial College London and Imperial College Healthcare NHS Trust has been awarded 1 million pounds (1.2 million euros) to conduct...

When is the best time to stop following low-risk prostate cancer?
There are a variety of factors associated with the discontinuation of active surveillance of men with low-risk prostate cancer, both demographic and clinical,...

ESR highlights MRI radiomics study of HCC
The European Society of Radiology (ESR) has promoted the findings of a European Radiology study that reported MRI radiomics features are associated...

MRI radiomics, AI distinguish GBM, brain metastasis
The combination of radiomics and a machine-learning algorithm can be highly accurate for distinguishing between glioblastoma (GBM) and brain metastasis,...