In 1987, Robert Getzenberg was beginning his doctorate at the Johns Hopkins School of Medicine urology department, the oldest in the country. Founded in 1915, the James Buchanan Brady Urological Institute’s historical roots seep from its every corner. Its current location boasts a cozy library showcasing the original chair and desk of the department’s founder. Exquisite medical illustrations—including the first to document prostate surgery—line the hallways. Here, under the eye of advisor Donald Coffey, a powerhouse in prostate cancer biology, the energetic young Getzenberg started hunting for distinctive molecular signatures of the disease in rat tumor tissue.

Prostate cancer biomarkers had recently become an enticing line of research. In 1986, a San Diego biotech, Hybritech, Inc., unveiled the first US Food and Drug Administration (FDA)-approved blood test for one such marker, called prostate specific antigen (PSA)—a protein that is leaked by damaged prostate cells, including cancer cells, into the blood.

But PSA was far from perfect. As the name implies, the protein is specific to the prostate, not to prostate cancer. Early tests found that most men with common (and benign) prostate inflammation also score high for PSA. So Coffey and Getzenberg were looking for a marker with fewer false positives.

For 50 years, microbiologists have known that the nucleus of a cancer cell looks drastically different from that of a normal cell: instead of forming a smooth circle, it typically has pinches in the membrane that make it look more like a lumpy snowman or a clover. In the 1970s, Coffey discovered the nuclear matrix—the three-dimensional mesh of proteins supporting a cell’s DNA—and suggested that this structure plays a part in the life cycle of the cell.

So that’s where the duo began their search, in rat models of prostate cancer tissue called ‘Dunning tumors’. Using gels that separate proteins on the basis of weight and charge, Coffey and Getzenberg and their colleagues found several proteins, including one called D-2, expressed in the Dunning tissue but not in controls.

In 1994, after finishing a two-year postdoctoral fellowship at Yale University, Getzenberg set up his own lab at the University of Pittsburgh. No one at Johns Hopkins had followed up on his Dunning tumor findings, so, with the blessing of his mentor Coffey, he seized upon the project anew. Within a few years, he says, he had identified D-2′s counterpart in human prostate cancer tissue, called early prostate cancer antigen-2 (EPCA-2).

In 2001, Getzenberg found a biotech company in Washington, Tessera Diagnostics, to invest in his biomarker work. Their goal was to produce a superior alternative to PSA testing, which by then bragged an annual global market value of about $300 million. (Beginning the following year, Tessera invested additional money in Getzenberg’s lab for investigations of biomarkers for colon and bladder cancer.)

In 2005, the then 39-year-old Getzenberg came back to Johns Hopkins, soon taking over for Coffey, who was still working but had stepped down as director of urology research. He moved into a modern remodel of Coffey’s magnificent former office, outfitted with a tank full of a dozen colorful fish and an even larger flat-screen monitor, and continued to work on development of blood assays for EPCA-2.

The research finally came to fruition in 2007, with a paper in Urology reporting significant new data. The study involved an analysis of blood samples stored at Johns Hopkins Hospital taken from 330 people: some who were being tested or treated for prostate cancer, some who had benign prostate conditions and some who had been diagnosed with various other cancer types. Getzenberg found elevated EPCA-2 cropped up in 94% of blood samples from men who were having surgery to remove prostate cancer. The marker was also specific: high EPCA-2 levels appeared in just 8% of the men without prostate cancer. What’s more, extremely high levels of EPCA-2 correlated with cancers that had spread beyond the prostate—a clear sign of aggressiveness.

Nevertheless, “we were pretty clear that we felt this was a proof-of-principle study, just to show that this marker could be found in the blood and that it could separate [people who have cancer from those who don't],” Getzenberg recalls from the large porch adjacent to his office.

That tempered message didn’t translate to the media, however. Johns Hopkins Medicine issued a press release, and dozens of popular news outlets picked up the story. Many of the articles quoted Getzenberg saying that large-scale clinical trials for the test could begin within nine months, putting it in doctors’ offices in one to three years.

That didn’t happen, and things soon turned sour with Getzenberg’s investors. On 11 February of this year, the biotech—now named Onconome—filed a lawsuit in Baltimore City Circuit Court against Getzenberg, the University of Pittsburgh and Johns Hopkins, alleging breach of contract and scientific fraud.

On 21 July, the judge dismissed the claims against the University of Pittsburg and Getzenberg, finding that the university was not subject to jurisdiction in Baltimore and that Onconome had agreed to pursue any claims against Getzenberg in Pennsylvania.

On 25 August, a month after obtaining new evidence from laboratory notebooks from Getzenberg’s lab, Onconome filed an amended complaint in Baltimore. In early September, it filed a similar case in federal court in Pittsburgh.

The new complaints state that the EPCA-2 test “was no more accurate in distinguishing cancerous tissue from normal tissue than flipping a coin” and that the company’s independent scientists could not replicate Getzenberg’s results. The company further alleges that Getzenberg’s lab technicians did not use blinded samples and that Getzenberg frequently presented “cherry-picked and unrepresentative selections” of his data to its board of directors.

The company claims to have lost $13 million as a consequence of Getzenberg’s alleged misrepresentations. When reached by Nature Medicine, Onconome’s founder and CEO, Raymond Cairncross, declined to comment on the lawsuit.

Likewise, Getzenberg says that institutional policy forbids him from commenting on pending litigation. But he eagerly defends the science behind EPCA-2, which he says is being shopped by other companies for commercial licensing.

The Pittsburgh case has not yet completed discovery, and the Baltimore judge has ordered a stay on the case until all parties attempt a mediation, which was scheduled for 4 December as Nature Medicine went to press. Whatever happens in court, there’s no denying that EPCA-2 development has been slower than the hype had predicted—a disappointment that seems to reflect a broader trend in the prostate cancer biomarker field.

“It’s a bit like waiting for the messiah,” says Gary Schwartz, scientific director of the Prostate Cancer Center of Excellence at Wake Forest University School of Medicine. “There’s been a lot of people who claim to be it, and a lot of hype about stuff, but, in my impression, it just hasn’t arrived.”

“That’s the natural history of biomarker research,” adds Shahrokh Shariat, instructor in urologic oncology at Memorial Sloan Kettering Cancer Center in New York. “A good biomarker has to be easier to measure, cheaper, faster and better than what we already have with PSA. And each of those criteria is really hard to reach.”

…read the rest of my latest feature in Nature Medicine