Liquid Biopsy

New device may track treatment response in ovarian cancer by analyzing usually discarded abdominal fluid

A microchip-based device developed by Harvard Medical School investigators at Massachusetts General Hospital may greatly simplify the monitoring of patients’ response to treatment for ovarian cancer—the most lethal form of gynecologic cancer—and certain other malignancies.

Ascites fluid from ovarian cancer patients is introduced through a filter into an inlet (left) where magnetically labeled benign cells are trapped. The remaining fluid passes through a microchip containing microwells where antibody-labeled tumor cells are captured (inset) for collection, imaging and additional analyses. Image: Hakho Lee and Jaehoon Chung, Mass General Center for Systems Biology

The team from the Mass General Cancer Center and the hospital’s Center for Systems Biology reported using the device to isolate and identify tumor cells from ascites, an accumulation of fluid in the abdomen that often occurs in abdominal cancers. The PNAS paper also describes developing a panel of four protein markers to accurately identify ovarian cancer cells in ascites.

“We were able to demonstrate that simply squirting small amounts of otherwise discarded ascites fluid into our device allowed us to quantify tumor cells and explore mechanistic markers of tumor progression without the need to process liters of ascites with advanced instrumentation not readily available in many community hospitals,” said Cesar Castro, HMS instructor in medicine at Mass General and co-lead author of the PNAS paper. “Moreover, achieving point-of-care readouts of tumor cell markers from repeatedly collected ascites at different time points could allow for frequent monitoring of treatment response without having to wait for the next imaging scan.”

The ability to track treatment response reliably lets caregivers know whether a particular anticancer drug should be continued or if another option should be tried. Tumor recurrence begins before metastases become visible on imaging studies, so several options for non-invasive “liquid biopsies” are being investigated, including analysis of circulating tumor cells and other factors found in the blood. Because ovarian cancer metastases are usually confined to the abdominal cavity and ascites commonly form in advanced disease, the research team theorized that ascites fluid could be an alternative, if not better, option than blood for treatment monitoring.

Isolation of ascites tumor cells has been challenging because they constitute less than 1 percent of the cells in ascites fluid. Ascites tumor cells themselves vary greatly in size, and other fluid contents—inflammatory and blood cells, cells from the abdominal lining and additional debris—often form large clumps that clog typical microfluidic devices. Along with removing the non-tumor-cell components of ascites fluid, the team also needed a way to accurately identify ovarian cancer cells and analyze their molecular properties.

Lengthy laboratory work compared ovarian cancer cells with benign cells and compared ascites samples from ovarian cancer patients with samples from individuals with noncancerous conditions like cirrhosis. That process led investigators to identify four protein markers that specifically identified ascites tumor cells from ovarian cancer patients. They confirmed the accuracy of the four-protein panel, called ATCDX, in two separate sample sets, comparing ascites fluid from ovarian cancer patients with either noncancerous fluid or with ascites from patients with other types of cancer.

Before the ascites fluid passes through the device, called the ATC chip, the sample is labeled with magnetic nanoparticles that bind to noncancerous inflammatory cells. The sample is introduced into the three-inch-long ATC chip through a filter that screens out clumps of debris. The sample then passes by a magnet that traps the magnetically labeled benign cells.

A mixture of antibodies to the ATCDX proteins, which label the markers for imaging detection, is also added to the device. After the magnetic sorting, the sample passes over a series of successively smaller microwells, which collect ascites tumor cells while even smaller leukocytes pass through the device. The concentration of ascites tumor cells captured on the chip is 1,000 times greater than it was in the original fluid sample.

The investigators initially tested their device by analyzing ascites samples collected from a single ovarian cancer patient over a 14-week course of treatment, first with standard chemotherapy and then with antiangiogenic therapy after disease progression resumed. The ATC chip revealed that the number of ascites tumor cells fell during initial treatment response, rose with tumor growth and fell again as antiangiogenesis treatment relieved the patient’s symptoms.

By analyzing the molecular properties of ascites tumor cells from 46 ovarian cancer patients, the researchers could distinguish between those whose tumors responded to treatment and those whose tumors did not respond.

“This device far exceeded our expectations,” said Ralph Weissleder, the Thrall Family Professor of Radiology at Mass General and senior author of the PNAS paper. “Coupled with our diagnostic panel, we were able to clearly distinguish between tumor cells and the extensive cellular debris commonly found in ascites. The ATC chip and the set of protein markers we uncovered, which reliably identified ovarian cancer cells floating in ascites, provide a novel platform for extending ascites tumor cells analysis to settings where the expensive equipment and labor-intensive techniques that ascites tumor cells isolation previously required would not be feasible.”

The research team noted that large-scale production of the ATC chip is already being planned. If future studies confirm their results, the device’s ease of use and low cost—estimated at less than $1 each—would make ascites tumor cells analysis a practical, valuable tool for both treating and researching ovarian cancer, and possibly other cancers whose tumors induce the formation of ascites, including pancreatic cancer.

The study was supported by National Institutes of Health grants P50-CA086355, R01-EB004626, R01-EB010011, P01-CA139980, U54-CA151884 and K12CA087723-11A1 and Department of Defense grant W81XWH-11-1-0706.

Adapted from a Mass General news release.