MR spectroscopy may be superior for determining prostate cancer prognosis

Detailed analysis of tissue chemistry could identify most appropriate treatment; more study needed

A new way of evaluating prostate tumors may help physicians determine the best treatment strategy. Using magnetic resonance (MR) spectroscopy, which provides detailed information on the chemical composition of tissue samples, researchers from Massachusetts General Hospital (MGH) have shown that chemical profiles of prostate tissue can determine a tumor’s prognosis better than standard pathological studies do. The report appears in the April 15 issue of Cancer Research.

“Our study indicates that analyzing prostate tissue’s metabolic profile may give clinicians additional information about the biologic status of the disease that could allow them, in consultation with their patients, to make better-informed decisions on the next steps to take,” says Leo L. Cheng, PhD, of the MGH Radiology and Pathology Departments, the report’s lead author.

Since the prostate-specific antigen (PSA) test became widely used to screen for prostate cancer, tumor detection rates have increased dramatically, particularly among those at early stages of the disease. But increased detection has led to a clinical dilemma, since standard histologic evaluation, based on a biopsy sample’s appearance under a microscope, often cannot distinguish which tumors are going to spread and which are not. Many men live for years with slow-growing prostate tumors before they die of unrelated causes, and treating such patients could cause more harm than benefit, Cheng notes. So finding a better way to determine which patients need aggressive treatment and which can try watchful waiting has been a major challenge.

Another problem is that a biopsy sample from one area the prostate may miss malignant cells elsewhere in the gland. Removal of the entire prostate can give a more definitive diagnosis, but if the tumor is a slow-growing one, the patient would have undergone unnecessary surgery. Surgery also is not appropriate when cancer has already spread beyond the prostate, since that situation requires other therapeutic approaches such a chemotherapy or drugs that block testosterone’s action.

Although MR spectroscopy has been used for many years to measure the chemical composition of materials, including biological samples, it has not been useful for analyzing tumor specimens. In recent years, Cheng and his colleagues have been developing a spectroscopic technique called high-resolution magic angle spinning that provides detailed analysis of a sample’s components without destroying its cellular structure. The current study was designed to evaluate the technique’s potential for providing information useful for clinical decision-making in prostate cancer.

The researchers used MR spectroscopy to analyze tissue samples from 82 patients in whom prostate cancer had been confirmed by prostatectomy. Almost 200 separate samples were studied, including many that appeared benign to standard histological examination. They then compared the spectroscopy results – detailed profiles of each sample’s chemical components – with the information gathered from pathological analyses of the removed glands and the patients’ clinical outcomes.

Several chemical components of the tissue samples were found to correlate with the tumors’ invasiveness and aggressiveness, supporting the potential of these metabolic profiles to provide valuable clinical information. Perhaps most significantly, even samples of apparently benign tissue had components that could successfully identify more and less aggressive tumors elsewhere in the prostate.

“Not only are the spectroscopy studies as good as histopathology in differentiating cancer cells from benign cells, they may be even better if they can find these metabolic differences in tissues that look benign,” says Cheng. “We need to do a larger scale, more systematic study of this technique before it can be applied to clinical practice. And we hope to collaborate with other institutions to identify different metabolic profiles that could provide additional information.” Cheng is an assistant professor of Radiology and Pathology at Harvard Medical School.