A new study documents a decade-long increase in the number of men who have incurable prostate cancer at their initial diagnosis, an ominous finding that prostate cancer-screening proponents have been predicting.

Both screening and diagnosis of early-stage prostate cancer have declined, coinciding with recommendations from an influential government advisory panel. In 2008, the U.S. Preventive Services Task Force said not to do routine PSA blood testing of men over age 74. And in 2012, it said to not screen any men - not even those at high risk - because the harms of unnecessary treatment outweigh the benefits of catching cancer early.

Despite less testing, death rates have been falling since 2007.

The new analysis, published in the journal Prostate Cancer and Prostatic Diseases, is the first to find an increase in metastatic cancers.

The Northwestern University analysis does not necessarily reflect a national trend. The study used a national cancer database from more than 1,000 hospitals rather than a representative sample of the population, so researchers could not calculate rates of disease.

Nevertheless, the statistics are stark: The annual number of metastatic prostate cancers at first diagnosis rose from 1,685 in 2004 to 2,890 in 2013 - a 72 percent increase the researchers call a "skyrocket" in a news release. The upturn was even greater among men 55 to 69, the age group believed to benefit most from early detection because of their risk level and life expectancy.

Over the same period, early-stage cancer diagnoses dropped from 25,708 to 16,223, or 37 percent.

Given that the surge in metastatic cases began before the U.S. Preventive Services Task Force's recommendations against routine PSA testing, other factors must also be at work, the Northwestern researchers concluded. They speculated that some tumors are more aggressive, and that better imaging technology has improved detection of the spread to lymph nodes or beyond.

But the implications, they wrote, "highlight the need for nationwide refinements in prostate-cancer screening."

"If you don't screen, you don't detect it" early, said senior author Edward M. Schaeffer, chair of urology at Northwestern's Feinberg School of Medicine. "This is going to affect mortality rates. Metastatic disease is incurable. They will die."

Eric Horwitz, a Fox Chase Cancer Center radiation oncologist, said the new study gives credence to anecdotal reports of spikes in metastatic cases - and to fears that the shift away from screening has gone too far.

"We really do need to adjust the guidelines," Horwitz said.

Thomas Jefferson University prostate cancer researcher Karen E. Knudsen said, "The challenge will be to refine guidelines so as to not miss men who would have benefited from earlier therapy, while at the same time not overtreating patients with a low likelihood of developing lethal prostate cancer."

In its controversial 2012 disavowal of PSA screening, the advisory panel concluded that early detection saves few, if any, lives, while treatment leaves many men with sexual or urinary problems or both. The panel said "most" cancers found by screening are innocuous and would not cause harm if left undiagnosed.

The task force, which influences insurance coverage, relied heavily on two major clinical trials that compared screening to no screening. A European trial showed screening reduced deaths by 21 percent. A U.S. trial found no difference in death rates, but it has been widely dismissed as flawed because so many men in the no-screening group got PSA tests anyway.

Asked about the new study, panel chair Kirsten Bibbins-Domingo, an epidemiologist at the University of California, San Francisco, said in an email: "The Task Force welcomes new research. . . . As part of our normal process, we are currently reviewing the latest evidence to update our 2012 recommendation."

She said that any man who is concerned about his risk of prostate cancer should talk to his doctor, "as this recommendation does not preclude a man choosing to be screened."

Debate about the value of screening has raged for decades for both prostate and breast cancers.

Even Dartmouth Medical School researcher H. Gilbert Welch, an outspoken critic of overdiagnosis and overtreatment, believes the advent of routine PSA testing a quarter-century ago explains a "steep" decline in metastatic prostate disease.

Welch analyzed a federal cancer database and found the incidence of metastatic cases fell by half in the seven years after the start of widespread PSA use in 1990. "It's hard to imagine another factor . . . exerting an effect so quickly," Welch wrote last year in the New England Journal of Medicine.

The Northwestern study's inability to calculate rates leaves unanswered questions. Perhaps the hospitals' increase in advanced disease simply reflected a proportionate increase in the vulnerable age group - namely, older men.

"It is very possible" that less early detection will result in more advanced disease, said Otis Brawley, chief medical officer for the American Cancer Society and a critic of overdiagnosis. "They simply did not prove it."

Prostate cancer remains the second-leading cause of cancer death among men, behind lung cancer, so finding the right balance between early detection and overdiagnosis is essential.

Schaeffer said his team plans to publish a profile of the metastatic patients in their study, analyzing their age, race, education, income, insurance coverage, and PSA level. The insights, he hopes, will help target screening to the men who need it most.

The task force four years ago "basically shot a volley across the bow, saying that screening the average person is not warranted," he said. "If you can understand who is diagnosed with metastatic disease, that would be one way to be smarter about screening."

215-854-2720 @repopter