CT texture analysis may aid risk stratification of renal masses

January 29, 2018

A recent study suggests experimental "texture analysis" of renal masses via computed tomography scan holds promise as a technique to allow more effective risk stratification.

San Diego-A recent study suggests experimental "texture analysis" of renal masses via computed tomography scan holds promise as a technique to allow more effective risk stratification. 

Use of the analysis technique appears to allow more detection of patterns "associated with clinical behavior and protein expression in renal cell carcinoma," said senior author Meghan Lubner, MD, of the University of Wisconsin School of Medicine and Public Health, Madison. The study findings were presented at the 2017 American College of Surgeons clinical congress in San Diego.

Read: Low-oxalate diet more beneficial than supplement

While CT and magnetic resonance imaging scans can identify large, solid renal masses, Dr. Lubner said, they offer no information about whether 75%-80% of small renal masses are cancerous. In addition, "For patients with small renal cell cancers," she said, "analysis of conventional imaging is unable to provide details for risk stratification such as tumor grade, RCC subtype, or identification of high-risk features."

Enter CT texture analysis, which offers an alternative to visual interpretation of renal masses, she said.

"CT texture evaluation is exciting because it allows computer quantification and analysis of statistical patterns within the image that may not be identified with visual inspection," she said. "We can see visually in many cases that a tumor is heterogeneous. It is just difficult to quantify in a meaningful way without a tool like this."

Next: How it works

 

According to Dr. Lubner, texture evaluation works this way: "We localize a renal tumor on CT by drawing a region of interest around it. A variety of software platforms are available, but basically the software collects data about the gray levels in the tumor."

This information includes details about brightness and how often the different gray level values occur, she said.

"Higher-order texture analysis may take into account the physical location of the pixels and their relationship to other pixels in the image."

For the study, the authors used the texture analysis technique to examine CT portal venous phase image scans of small renal cancers (≤4 cm) in 249 patients (105 women, 144 men, average age=56.6 years).

Among other findings, the authors linked texture "features" to histologic cell type (clear vs. non-clear, p<.001).

"Pixel patterns that are identified statistically are associated with clinical behavior and protein expression in renal cell carcinoma," Dr. Lubner said. "In general, more heterogeneous tumors are more aggressive, and certain texture features may predict histopathologic features including whether the tumor is clear cell versus non-clear cell, but also may be associated with additional tissue biomarkers; for example, proliferative index or markers of tumor neovascularity."

Also see: Is favorable-risk GG2 prostate Ca suitable for active surveillance?

Dr. Lubner cautioned that, "Currently, this technique is being used for research purposes only and is not ready for routine clinical use. While there is promising clinical data emerging, there are many challenges and unknowns around texture analysis that need to be evaluated and overcome prior to widespread or mainstream use."

Going forward, she said, "More pathologic correlation is needed to aid in understanding what exactly these texture features are representing and their causal relationships to the tumor biology. It is unlikely that this technique would supplant histopathologic sampling in the short term, but it could suggest that aggressive features may be present and push for a biopsy-or repeat biopsy if initial biopsy doesn’t support this-rather than surveillance."

In addition, texture analysis "could corroborate biopsy results. It could be useful in discussions of treatment and prognosis if aggressive features are present."

Other researchers are exploring the technique in prostate cancer, she said. And it's being examined in renal cell cancer patients receiving chemotherapy as a tool to monitor their response to treatment, she said.

Dr. Lubner reports grant funding from Philips and Ethicon. Other study authors report grant funding from NIH/National Cancer Institute and various consulting and shareholder disclosures. One co-author is co-founder of VirtuoCTC.

More from Urology Times:

How to manage your reputation, recruit patients online

Tax law may threaten Medicare, provider pay

How important is urology/gynecology collaboration?

To get weekly news from the leading news source for urologists, subscribe to the Urology Times eNews.