Early detection – we cannot stress this enough when it comes to cancer. In the U.S.A., if breast cancer is detected before spreading out of the breast (stage I) the 5-year survival rate is 99% but if it is not diagnosed until it has become metastatic (stage III) the rate is 72%. This trend is true for cervical cancer as well. Strikingly, if colorectal cancer (CRC) is detected at the precancerous adenoma stage the disease can be prevented, and early, local CRC has a 90% 5-year survival rate that drops to 70-12% if the tumor has spread to other tissues. We highlight these three cancers because physicians have screening tools that allow early detection: mammography for breast cancer, pap smear tests for cervical cancer, and a combination of colonoscopy and fecal tests for CRC. All of these screening tools have proven useful and increased the early detection of cancers; however, there is plenty of room to improve. In the case of CRC, colonoscopy provides the greatest sensitivity for detecting advanced adenomas and early stage tumors, but is invasive, expensive, and time-consuming for patient and doctor alike resulting in low rates of screening. Current fecal tests have low sensitivity for adenomas and are being used less frequently each year. The ideal test for early detection of CRC (and many other cancers) would be finding disease specific markers in patient serum, the so called liquid biopsy. Annual blood marker panels are a common practice for patient groups at risk for CRC, making adherence very easy. Dr. Paul Lampe’s Laboratory (Public Health Sciences and Human Biology Divisions) at Fred Hutch specializes in finding cancer-specific signatures in blood samples to serve as diagnostic biomarkers for detecting disease. While the lab is interested in finding biomarker panels for multiple cancers, Dr. Jon Ladd is focusing on early detection of CRC and recently published results in Gut this month.
Figure provided by Dr. Jon Ladd
For this study researchers measured protein levels within subject plasma and looked for specific proteins that were elevated in those with CRC. Other groups have taken this approach and identified biomarker panels for this disease, but like current fecal tests these panels had low sensitivity for adenomas and early stage tumors. To improve early detection researchers used subject samples originally collected from a screening study to understand risks of heart disease. This group contained participants that never developed CRC and those that were diagnosed 0-1, 1-2, or 3 years after the blood sample was collected. This retrospective approach served as an ideal model for early detection. From this pool of samples researchers identified two groups of 79 subjects each, those that were diagnosed with CRC and those without CRC. Protein content in samples was measured with a custom antibody array containing ~3200 different antibodies selected by the researchers. From these arrays 78 markers were found to be elevated specifically in subjects who subsequently developed CRC.
These markers were further validated using samples from the Early Detection Research Network (EDRN). This collaborative project allowed researchers to analyze 30 samples from each group of subjects with specific CRC pathologies (control, adenoma, advanced adenoma, early (stage I-II) or late (stage III-IV) CRC). While most of the 78 biomarkers identified were elevated in the late stage CRC group, 32 protein markers were elevated in samples from all stages of CRC. From these, five were chosen as the best performing: BAG4, IL6ST, VWF, EGFR, and CD44.
The increased plasma level of these five proteins was also observed by immunoblotting, with BAG4 and IL6ST showing the most robust increase. BAG4, IL6ST, and CD44 also showed increased expression within tumors when assayed by immunohistochemistry. Interestingly, some markers (CD44, VWF, and EGFR) were elevated in response to other cancers (breast, lung, and pancreas). Dr. Ladd explained the value of looking at all markers together as a disease-specific panel, “A ‘pan-cancer’ test might not be as useful as multiple, disease-specific tests. If the test was positive, there would be too many options for follow-up. Some biomarkers that are common among cancers could, however, be useful in combination with other markers that would confer disease specificity to the test. From our study, BAG4 and IL6ST appear to confer that specificity for colon cancer.” Thus CD44, VWF, and EGFR may serve as a ‘cancer’ baseline to pair with disease specific markers.
Based on these confirmations, researchers developed a high-throughput Luminex assay for the five protein markers and screened a larger sample set. Subject blood samples were collected at Ogaki Municipal Hospital in Japan prior to diagnostic colonoscopy. Thus researchers were able to expand to 168 subjects with no evidence of colon cancer, 159 subjects with low-risk polyps, and 514 positive for CRC. By analyzing all five proteins together, this assay had 73% sensitivity for identifying patients with CRC, while current fecal tests range from 23-82% sensitivity. Early detection of adenoma with this panel performed better than current tests (53-55% versus 18-24% sensitivity respectively). These findings indicate this plasma protein biomarker panel is equivalent or better to current FDA-approved tests. However, these findings must be scaled up to even larger sample sizes, said Dr. Ladd “The next step for us is to further test the specificity and utility of these markers in larger cohorts. We are currently evaluating our markers in ~4700 samples to further validate their utility in diagnosing pre-cancerous adenomas, examine their specificity to colon cancer, and evaluate the effect of co-morbidities, e.g. diabetes, heart disease, etc., on marker performance.”
Rho JH, Ladd JJ, Li CI, Potter JD, Zhang Y, Shelley D, Shibata D, Coppola D, Yamada H, Toyoda H, Tada T, Kumada T, Brenner DE, Hanash SM, Lampe PD. 2016. Protein and glycomic plasma markers for early detection of adenoma and colon cancer. Gut. Published Online First: 07 November 2016.
Funding for this research was provided by the Cancer Research Institute, the Kitsche Family Memorial, and the Geriatric Research Education and Clinical Center.