Unique protein expression could serve as a screening test for head and neck cancers

May 10 2006

The blood of patients with head and neck cancer appears to have unique patterns of protein expression that one day could serve as a screening test for the highly aggressive cancer that is often diagnosed too late, researchers say.

Studies comparing protein expression in 78 patients with head and neck cancer to 68 healthy controls revealed numerous differences in protein expression, Medical College of Georgia researchers say.

"We found scores and scores of proteins that were differentially expressed," says Dr. Christine Gourin, MCG otolaryngologist specializing in head and neck cancer and the study's lead author. "We found there are at least eight proteins whose expression significantly differs between controls and people with cancer."

This protein fingerprint correctly classified study participants as cancer patients with a high degree of sensitivity and specificity - 82 percent and 76 percent, respectively, according to research published in the current issue of Archives of Otolaryngology.

"If these results hold up over time, they would suggest that this would be a good screening test for at-risk people," Dr. Gourin says. "Right now there is no good, effective screening test for head and neck cancer short of physical examination. Unfortunately it takes the development of symptoms to warrant a visit to the doctor, such as a sore throat; ear, tongue or mouth pain; painful eating or swallowing; or a change in the voice. Sometimes the first sign is a lump in the neck which is already a sign of an advanced tumor that has spread to the lymph nodes."

Belated diagnoses translate to fairly dismal survival rates: less than 50 percent five years following diagnosis of stage three or four tumors, Dr. Gourin says. The rare patient who is diagnosed early faces much better odds: voice box cancer caught in stage one has about a 95 percent five-year survival, for example. The goal is to screen high-risk populations - those with a history of alcohol and/or tobacco use - as well as those with head- and neck-specific complaints who don't have those risk factors, says Dr. Gourin. She notes that about 20 percent of head and neck cancer patients have no history of alcohol or tobacco use.

Advanced proteomics technology - which can be applied to many tumor types - enables protein expression to be plotted on graphs that illustrate peaks and valleys. "Sometimes the underexpression of a protein may be significant," Dr. Gourin says.

The unique patterns surfacing may one day provide more than screening. Study findings indicate the protein fingerprint also is highly successful at classifying specific types of head and neck cancer, correctly classifying 83 percent of oral cavity tumors and 88 percent of laryngeal tumors, as examples, researchers say.

Also, in a small subset of 12 patients, protein expression helped researchers correctly classify how cancers responded to treatment, indicating its effectiveness in long-term follow-up, Dr. Gourin says. "We could easily use this to follow patients for life and detect any recurrence early as well as improve our ability to detect a second primary tumor, which occurs in about 8 percent of people," she says.

Clinical availability of a screening test for head and neck cancer is still years away, says Dr. Bao-Ling Adam, MCG cancer researcher and study co-author. But the researchers are continuing to make progress, already collecting more patient data to ensure that the patterns they have identified in the blood are effective biomarkers for head and neck cancer. Dr. Gourin is considering opening the study to other medical centers to increase numbers possibly into the thousands.

They also want to know if protein expression patterns found in the blood are expressed by cancer cells themselves, says Dr. Adam, who has begun doing proteomics studies on the cancerous tissue of surgery patients to find out. "What we see in the blood could be from the cancer cells or from the body's response to cancer," she says.

If they are the same, the proteins also could yield novel therapeutic agents, Dr. Adam says.

This will help solidify the link between the protein patterns and cancer as well. "For screening you really have to use body fluids: blood, saliva, urine," says Dr. Adam. "When the normal cell transforms to a cancer cell, we want to see the changes within the cells. When we find the protein differences between cancer cells and normal cells, we can use this information to detect head and neck cancer."

Interestingly, to date they have not found any proteins expressed by cancer that are not expressed normally; the difference is a matter of degrees of expression, says Dr. Adam, who also is using proteomics to find a better biomarker than prostate specific antigen, or PSA, for prostate cancer.