By Clinical Research News Staff 

June 16, 2026 | For the roughly 1.3 to 2 million women who visit their doctor each year with abnormal postmenopausal bleeding, the diagnostic journey has long been an uncomfortable one. The standard workup typically involves an endometrial biopsy—an invasive, often painful procedure that, despite its widespread use, fails to yield a conclusive diagnosis in more than 30% of cases. 

Two veteran diagnostics executives think there's a better way, and they're betting nearly $45 million in venture funding on it. 

PinkDx cofounders Bonnie Anderson and Giulia Kennedy, Ph.D., have spent the last four years building a vaginal swab test that can detect — or just as importantly, rule out — endometrial cancer using whole-transcriptome RNA sequencing and machine learning. The pair previously worked together at Veracyte, the genomic diagnostics company Anderson co-founded in 2008, where they developed classifiers that reduced the need for invasive surgeries across multiple disease areas. Now they're applying the same playbook to gynecologic cancers. 

Their first major clinical milestone came from a 236-patient exploratory study conducted at Mayo Clinic, Columbia University, and Houston Methodist, which confirmed that cancer-associated molecular signals could be captured from a simple vaginal swab without ever accessing the uterus. That finding, published in the International Journal of Gynecological Cancer, set the stage for what comes next. 

That next step is now underway. The PROACTION study, a prospective, multi-center trial enrolling close to 3,000 women across 16 sites, aims to validate those early findings in the broader, more diverse populations that will ultimately use the test once it reaches the market. Sites range from Meharry Medical College and the University of Alabama Birmingham to Baylor College of Medicine and Indiana University, a deliberate effort to address a disease that disproportionately affects Black women, who die from endometrial cancer at twice the rate of white women. 

The trial is enrolling women aged 45 and older presenting with abnormal uterine or postmenopausal bleeding, exactly the patients the test is designed to serve. Database lock is expected within six months, with follow-up data collection continuing at three and five years. 

Kennedy, who serves as chief scientific officer, has engineered the test's machine learning models with a specific clinical priority in mind: sensitivity. When the goal is ruling cancer out, missing a positive case carries far greater consequences than a false alarm. "The models are chosen, selected, modified, tuned and developed over and over to be more and more sensitive," she says. 

PinkDx expects to validate its first product by the end of this year and launch by mid-2027, initially as a laboratory developed test while the company builds the evidence base needed for insurance coverage. 

For Deborah Borfitz’s complete story, visit Diagnostics World.