Machine Learning Uses Lung Cancer Scans to Predict Heart Damage

As patients with lung cancer live longer, the risk of long-term cardiac side effects of radiation therapy has been increasing, despite advances that reduce the radiation dose to the heart. New research uses machine learning to mine data from standard lung-cancer scans and predict patients most likely to have heart damage from radiation treatment later in life.  If confirmed by future tests, this approach could identify at-risk patients and ensure they are appropriately monitored before their condition becomes debilitating.

Along with chemotherapy, radiation therapy remains the definitive and standard of care for advanced-stage lung cancer. During treatment, patients routinely undergo PET/CT scans to monitor the cancer over time. In a recent study, Jefferson researcher and first author Wookjin Choi, PhD, developed an artificial-intelligence algorithm that could re-analyze these PET/CT images and accurately predict patients with more inflammation in the heart – often a precursor of long-term damage.

“This is the first application of functional radiomics – or using algorithms to extract quantifiable information from medical images – to predict future cardiac toxicity,” says senior author Yevgeniy Vinogradskiy, PhD.

The researchers were careful to counteract some of the common pitfalls of artificial intelligence, such as bias and lack of clarity in how it makes conclusions. First, they trained the algorithm on data from three sources to diversify the patient images and reduce the likelihood of bias. Second, rather than use pre-written or ‘black-box’ algorithms, Dr. Choi wrote his own code, and winnowed the predictive features down to nine characteristics. “Nine is a much more manageable number to troubleshoot, and because we know which features we’re looking at, we have confidence that they’ll be clinically meaningful,” he says.

“What’s exciting about this approach,” says radiation oncology department chair Adam Dicker, MD, PhD, “is that it could provide important additional diagnostic information with no additional scans, sparing the patient extra costs and radiation exposure.”

Edyta Zielinska

Link between Prenatal Drug Use and ADHD

Prenatal use of alcohol, tobacco and cannabis have all been independently associated with adverse health impacts for the baby. But many people who use these substances during pregnancy aren’t using them in isolation. “What we know from people’s habits is that most people are co-using substances,” says Jennie Ryan, PhD, CPNP-AC, a nursing researcher at Thomas Jefferson University.

In a new study in the American Journal of Preventive Medicine examining how patterns of prenatal substance use impact attention-deficit/hyperactivity disorder (ADHD) diagnoses in children, researchers found that co-use of alcohol and tobacco, and use of cannabis alone, posed the greatest risk.

The study examined prenatal alcohol, tobacco and cannabis use alone and in combination with each other, says Dr. Ryan, the lead author. The strongest association was with alcohol and tobacco co-use during pregnancy—about four times greater odds of parent-reported ADHD in the child. “These are two of the most readily available substances,” Dr. Ryan says, “so it’s important to educate pregnant people about this risk.”

The next strongest association was with cannabis-only use, about two times greater odds of parent-reported ADHD. Surprisingly, Dr. Ryan says, cannabis combined with other substances did not pose as strong an ADHD risk.

The research data came from the Adolescent Brain Cognitive Development (ABCD) Study, a longitudinal repository of almost 12,000 diverse children and families across the United States. But while the ABCD Study is a large and rich data set, Dr. Ryan says it does not include ADHD diagnoses for parents. This is a notable limitation because there is a genetic component to ADHD.

The next step? Dr. Ryan is now using the ABCD Study’s longitudinal data to examine how the treatment of ADHD in children, such as medications or cognitive behavioral therapy, impacts their risk of developing substance use disorder later in life.

By Christina Hernandez Sherwood

Understanding Potassium Channels to Design Better Drugs

Potassium channels are openings that allow charged potassium atoms to cross the cell membrane. Voltage-gated potassium channels — which open only when a specific voltage is reached across the cell membrane — are essential for the electrical impulses that nerve cells or neurons use to communicate. Dysfunction of these channels is implicated in diseases like epilepsy, autism and schizophrenia.

To treat these conditions, certain voltage-gated potassium channels have become targets for drug developers. Over a decade ago, researchers at Autifony Therapeutics synthesized a molecule called AUT5, which helps certain voltage-gated potassium channels stay open. But its mechanism of action was unknown, making it difficult to develop AUT5 into an effective, specific, and safe drug.

To learn how AUT5 binds to potassium channels, neuroscientist Manuel Covarrubias, MD, PhD and his research assistant Qiansheng Liang, MD led a study published in Nature Communications that focused on Kv3 channels, the subclass of voltage-gated potassium channels that selectively bind AUT5. They are connected to multiple neurological diseases.

Through a multi-prong process involving electrophysiology, structural biology and cryo-electron microscopy, a technique that allows direct visualization of membrane proteins at near-atomic resolution, Dr. Covarrubias’ lab and global collaborators identified a “pocket” in Kv3 as the binding site for AUT5. This pocket is situated between the channel’s voltage sensor and the opening that potassium travels through, and next to a region known as the “extracellular turret”, named for its resemblance to a tower projecting above a castle. The structure of the extracellular turret is such that it helps secure the binding of AUT5, like a well-fitting glove.

“When we started, we knew what AUT5, the ‘hand’, looked like, but we didn’t know what or where the ‘glove’, the binding site, was,” says Dr. Covarrubias. “Now, we know the shape of the glove, and how it wrinkles to keep the Kv3 pore open.”

This research will help drug companies develop even more effective compounds to treat potassium channel-related diseases. Clinical trials using AUT5 and closely related compounds are already underway to treat children with intractable epilepsy and fragile X syndrome.

By Marilyn Perkins