If future studies show their test -- measuring the level of activity of a signaling pathway called Hedgehog -- can predict which prostate cancers will spread, the results could revolutionize decision making processes for prostate cancer patients, the researchers say.

Most prostate cancers grow slowly, making "watchful waiting" a common alternative to immediate surgical removal of the prostate. However, there's no sure-fire way to tell whose cancer will stay put in the gland, and whose will be aggressive and spread -- a development that despite aggressive treatment is usually fatal.

In the September 12 advance online edition of Nature, the Hopkins researchers report that only three of 12 localized prostate tumors obtained at surgery had detectable activity of the Hedgehog signaling pathway. In contrast, all 15 samples of metastatic prostate cancers, donated at patients' deaths, had Hedgehog activity, which was 10 to 100 times higher than the highest levels seen in localized tumors. It remains to be seen whether Hedgehog activity in localized cancers will predict the ability to be metastatic.

The Hedgehog pathway produces a well-known growth and development signal during embryonic and fetal stages. It is also active in some cancers, including prostate, pancreatic and stomach cancers and the brain tumor medulloblastoma, but the researchers' study is believed to provide the first evidence of its role in cancer's spread.

"If we can use Hedgehog activity to predict whether a tumor will metastasize, we will have a great diagnostic tool, but manipulating the Hedgehog signaling pathway may also offer a completely new way to treat metastatic prostate cancer," says David Berman, M.D., Ph.D., assistant professor of pathology, urology and oncology at Johns Hopkins. "Right now nothing works very well -- you can help temporarily by cutting off testosterone, but the cancer always comes back."

In experiments with mice, fellow Sunil Kahadkar, M.D., showed that blocking the Hedgehog signal with daily injections of either a natural plant compound called cyclopamine or an antibody slowed and even reversed growth of highly aggressive rat prostate tumors implanted into the animals. Without treatment, the aggressive cancers, from a collection established by Hopkins' John Isaacs, Ph.D., killed the animals within 18 days. A low dose of cyclopamine gave the animals an extra week to 10 days, but at a higher dose, these aggressive cancers not only didn't metastasize, they actually disappeared and didn't return.

In a similar set of experiments using human prostate cancers implanted into mice, treatment with cyclopamine also caused those tumors to regress and not return -- even months after treatment was stopped, the researchers report.

"Cyclopamine may not itself become an anti-cancer drug, in part because it's already in the public domain -- it's been known since the mid 1960s as the cause of one-eyed sheep in the western U.S.," says Philip Beachy, Ph.D., professor of molecular biology and genetics in Hopkins' Institute for Basic Biomedical Sciences and a Howard Hughes Medical Institute investigator. "But our finding that cyclopamine inhibits Hedgehog signaling has provided the basis for drug companies' very active efforts to develop new mimics of cyclopamine."

Right now, prostate cancer is evaluated largely by levels of prostate specific antigen (PSA) circulating in the blood. However, the ranges associated with various potential diagnoses -- non-cancerous growth, cancer, and aggressive cancer -- are fairly rough guides. And even under a microscope, aggressive prostate cancer doesn't always look appreciably different from its wallflower counterpart.

In sharp contrast, levels of Hedgehog activity weren't even close between still-localized tumors removed during prostatectomies and those from lethal metastatic prostate cancers, which were collected as part of a research program run by G. Steven Bova, M.D., assistant professor of pathology, to try to figure out what makes them so deadly.

To investigate Hedgehog's role in metastasis, Karhadkar genetically engineered normal prostate cells to activate their Hedgehog signal. These cells then grew unchecked and formed aggressive tumors when implanted into mice, he found. He also discovered that triggering Hedgehog activity in a low-metastasizing rat prostate cancer line made it metastasize aggressively.

"Hedgehog isn't just making these cells grow and divide more, the signal is really converting them from being indolent to being highly invasive and dangerous," says Beachy.

Exactly how the Hedgehog signal is involved in other cancers, including pancreatic and stomach cancers and medulloblastoma, a childhood brain cancer, is still being worked out. Critical in normal embryonic development, the signal is supposed to be turned off when cells take on the "grown-up" identity of a differentiated cell type.

Karhadkar, Beachy and Berman -- and a growing number of other scientists -- point to the involvement in cancer of embryonic proteins and pathways like Hedgehog as evidence that aggressive cancer in particular might form not by accumulation of genetic errors in regular cells, but because a smaller number of errors occurs in a more primitive cell, what might be called a "stem cell," in the tissues. And it would be these "cancer stem cells" -- transformed versions of the tissue's normal stem cells -- that metastasize and travel through the body to form new tumors in distant places.

"Perhaps aggressive prostate cancers get started from a more primitive prostate cell or from a different initiating lesion than do prostate cancers that don't metastasize," says Beachy. "It's an idea we're exploring."

The research was funded by the National Institutes of Health, the Prostate Cancer Foundation, and the Howard Hughes Medical Institute. Authors on the paper are Karhadkar, Bova, Nadia Abdallah, Surajit Dhara, Anirban Maitra, Isaacs, Berman and Beachy, all of Johns Hopkins; and Dale Gardner of the U.S. Department of Agriculture's Poisonous Plant Research Laboratory.

Under a licensing agreement between Curis Inc. and the Johns Hopkins University, Beachy and Berman are entitled to a share of royalty received by the University on sales of products described in this article. Beachy and the University own Curis Inc. stock, which is subject to certain restrictions under University policy. Beachy and Berman are paid consultants to Curis Inc. and Berman is a paid consultant to Genentech Inc. The terms of this arrangement are being managed by the Johns Hopkins University in accordance with its conflict of interest policies.