Magnetic resonance spectroscopy-based method may help predict prostate cancer aggressiveness

Metabolic analysis of benign tissue samples from men with prostate cancer allowed researchers to retrospectively predict the aggressiveness of disease and likelihood of recurrence, according to study results.

Leo L. Cheng, PhD, associate professor of radiology at Harvard Medical School, and colleagues suggested that prostate cancer alters cellular metabolism before formation of the malignancy visible by histology.

Magnetic resonance spectroscopy-based metabolomics could help determine whether a patient should have a prostatectomy or undergo active monitoring.

Cheng and colleagues conducted a retrospective analysis in which they assessed 365 histologically benign tissue samples from men with prostate cancer. Metabolic profiling of the samples demonstrated elevations in myo-inositol among men with aggressive disease.

The researchers said their approach also could make it possible to identify patients with less aggressive prostate cancer at the biopsy stage and determine recurrence risk.

H emOnc Today spoke with Cheng about the discovery, what is required to validate the results, and how the findings could lead to cost savings in the health care system.

Question: Can you provide some background about how you made this discovery?

Answer: We have been studying human prostate cancer tissue for the past 15 years. The reason we got into this area is because we invented a methodology of magnetic resonance spectroscopy that allows us to look at the cellular tissue chemistry without destroying the tissue. It is very important for this study that tissue not be destroyed; this way, histology can still be done. The samples we looked at were from surgeries performed about 10 years ago. Because of this new capability of measuring pathology after spectroscopy analysis, we were able to evaluate the cellular chemistry according to the tissue pathology. Ninety percent of the tissue we evaluated was histologically benign tissue from cancerous prostates. Our findings agree with the current successful rate of random biopsy for detecting cancer in prostate initiated by PSA testing. Prostate biopsy procedures used to take six cores, but that was not good enough to catch cancer lesions. Now they take 12 cores. As for the tissues we sampled, we got them from patients who underwent surgery. We knew they were from cancerous prostates, but we didn’t know if those tissues were actually cancerous. With pathology information evaluated, we could analyze spectroscopy data from histologically benign tissue for correlations with patient clinical status, including pathological status, which now can only be measured from the whole removed prostate. From a benign tissue, it is difficult for pathology to say anything about the status of the whole prostate, but tissue metabolomics may.

There are two ways to consider this information. One is that, histologically, it is still benign, but it has changed biochemically. Another possibility is that metabolites can move from cancer lesion to the surroundings to give us correlation. With benign tissue, we measure the metabolic profile using statistical tools to collectively consider all the measurable metabolites. With this approach, we can predict chemical recurrence of prostate cancer. That is important because we can achieve all these measurements with small piece of prostate tissue, even if it is benign. This may give us information to predict what the outcome might be after treatment to guide treatment strategies.

Q : Are efforts underway to validate the findings in a larger sample?

A : Yes and no. This is the first study to propose this idea of using benign tissue to predict pathology and patient outcomes. There haven’t been any multicenter studies to validate this yet. However, we are actively working on this. Over the past 5 years, we have collected data from patients who underwent prostate biopsy. The ones we published were from patients who underwent prostatectomy. About 5 years ago, we started actively collecting metabolomic information from those who underwent biopsy. They visited multiple urologists and, each time they got a biopsy, we got one or two biopsy cores. We analyzed the chemistry and gave it back to pathology, so it was still part of patient clinical data. This made it almost like a prospective study.

Q : Are data from th at study available?

A : We are in the process of writing it. We are reporting on fusion biopsy in meetings. Our urology colleagues went to an American Urology Association meeting and presented this paper. They received an award for the poster. A couple days later, at the Annual Meeting of the International Society of Magnetic Resonance in Medicine, our paper was selected as plenary talk. It seems people are interested in this. So we are moving along, but we haven’t specifically verified this paper yet.

Q : Is it too early to talk about this as viable strategy, or one that could be practice - changing?

A : It’s not too early. The most important thing for prostate cancer is not early diagnosis. We have many cases we can diagnose early. The issue is the implication of cancer to patients. When you look at old data, before PSA testing, you see men aged older than 80 years had cancer in their prostates but died of something else. But at age 50 years, if you get diagnosed with prostate cancer without knowing its malignant potential, you will probably do something about it. It’s not benign if you get a prostatectomy, because this is a surgical procedure. It changes your life. If we can identify cancers that are likely to kill you, we can treat them, but we also can monitor those cancers that aren’t likely to kill someone. We won’t be able to predict this for all patients, but maybe for a subgroup of 20%. If you can identify a subpopulation of that size, it is clinically meaningful. The old paradigm was to eradicate prostate cancer completely. Now the question is, how do we live with it or treat it as a chronic condition? This may be able to give patients confidence to live with cancer.

Q : Thinking beyond the clinic, how can this approach lead to more active surveillance and/or watchful waiting, and, consequently, cost savings to the health care system overall?

A : If you go to do a prostatectomy, you can find out how much it will cost in different hospitals in different areas. If you get 20% of people who you can confidently tell that they don’t need a prostatectomy — in other words, they will benefit from watchful waiting or active surveillance — that is a huge amount of money. –by Rob Volansky

Reference:

Vandergrift LA, et al. Sci Rep. 2018;doi:10.1038/s41598-018-23177-w.

For more information:

Leo L. Cheng, PhD, can be reached at Athinoula A. Martinos Center for Biomedical Imaging, 149 13th St., Charlestown, MA 02129; email: lcheng@mgh.harvard.edu.

Disclosure: Cheng reports no relevant financial disclosures.