Blood in the urine is a worrisome surprise, even though it's usuallypainless. When urologists examine patients with hematuria, theythink kidney stones, urinary tract infection or bladder cancer.

With 50,000 new cases each year, carcinoma of the urinary bladder isthe fourth commonest malignancy among American men, and theeighth among women. But it doesn't rank high as a killer.

"Mortality is directly related to the stage of the disease," said clinicalurologic oncologist Mark Schoenberg, of Johns Hopkins University,in Baltimore. "Death from superficial _ early stage _ bladdercancer, which is 75 percent of all cases, is negligible. Only 10 percentof those people," he told BioWorld Today, "go on to developinvasive disease, and half of them die from bladder cancer over aperiod of five-year follow-up."

That follow-up, though, as well as initial diagnosis, can be invasive,costly and cumbersome, Schoenberg observed. Besides X-rayimaging and exfoliative cytology (biological examination of bladder-wall cells shed into the urine), it involves cystoscopy _ "looking intothe bladder and urethra with an endoscope." The flexible, opticalexploratory cytoscopy tube incorporates mini-instruments fordissecting and removing tumors visualized on the bladder's innerwall.

However, patients thus relieved of their initial tumor burden are nothome free.

"About 70 percent of these little tumors will recur," observedHopkins medical oncologist David Sidransky, "and over time theybecome large. So such patients get repeat cystoscopy every three tosix months, trying to see that they don't acquire a big tumor."

Sidransky told BioWorld Today that he is "trying to develop a testthat in these situations is simple and non-invasive."

In today's Science, he, Schoenberg and their co-authors report"Molecular detection of primary bladder cancer by microsatelliteanalysis." Sidransky, the paper's senior author, explained thisstrategy: "It's comparing a patient's urine DNA to his or her bloodDNA by DNA amplification via polymerase chain reaction [PCR]."

Microsatellites Finger Tumor Cells

This method, as the team reported in Science, correctly diagnosedcancer in the shed bladder cells of 19 out of 20 patients (95 percent)with gross hematuria. Another five non-symptomatic controls provednegative. Conventional cytology scored nine of 18 subjects, (50percent), which, Schoenberg said, is typical of that technique.

Central to Sidransky's molecular strategy is spotting microsatellitesin the PCR-amplified urinary DNA of shed bladder cells. Theserepetitious stuttering sequences are the tip-off that such cells arecancerous.

"Cancer cells are clonal," he observed. "If they make a mistake in theform of a microsatellite, they will transmit that deletion or expansionof certain base pairs to their daughter cells. And we can identifythese," he continued, using PCR, "simply by comparing urine DNAto normal DNA in the patient's blood."

The notorious triplet base-pair repeats that hex Huntington's disease(HD) and other neurodegenerative disorders, Sidransky said, "are apart of the disease gene. Two-thirds of all our repeats are in junkDNA, among genes, not inside genes." (See BioWorld Today, Aug.28, 1995, p. 1.)

He continued: "Most of the microsatellites that we tested were totallyanonymous, just spread out in the genome. We don't specifically pickthem because they're next to one gene or another. They simply occur,and are clonally propagated by all daughter cells."

Another difference, he pointed out: The triplet that HD repeats andexpands, CAG, "affects the function of the HD gene. We usemicrosatellites as tumor markers, in the sense that they're telling usthere is clonal proliferation."

Mom-Or-Pop Chromosome Gap Loses p53 Too

Microsatellites are only half the story. Equally informative isdetecting the deletion of a single chromosome from the genomicDNA of a suspect urinary bladder cell.

"We can find these chromosomal deletions with the samemicrosatellite markers," Sidransky said. "Every person has twocopies of every non-germline chromosome," he explained, "one frommom, one from dad. And when we compare them to blood, what yousee in our tumors is not a new marker by base-pair expansion, but theabsence of a marker. And this," he added, "was a completelyunexpected finding."

The deletion of those mom-or-pop chromosomes, Sidranskysuggested, "is knocking out very important genes involved in theprogression of cancer. For example, loss of one human chromosome17 is in the area where the tumor suppressor gene p53 is located, andin our tumor cells p53 is gone."

He observed, "we can now use this information in this type ofdiagnostic strategy, which is what we did." The test he designed andproved in the pilot study, Sidransky said, "now leaves my lab andgoes to the clinical setting, where we hope we can automate it in afew months."

He and co-author Schoenberg are planning a larger, prospectivestudy, involving three or four other clinical centers, Schoenberg said,"as well as identifying several hundred patients at high risk ofdeveloping bladder cancer, who would be suitable for some type ofscreening."

He said the test should be available "in a research setting in the verynear future, but in terms of its availability as a general diagnostic testthat one could order in the doctor's office, that certainly is yearsaway."

As for cost, the Science paper stated, "in principle, it can beperformed at about one-third the cost of cytology."

Sidransky's head-and-neck-cancer laboratory at Hopkins is funded byOncor Inc., of Gaithersburg, Md., which holds exclusive worldwiderights to his three pending patents, Sidransky said, which are assignedto the university. "They cover general molecular detection strategiesfor all types of cancer." (See BioWorld Today, Oct. 13, 1994, p. 2.)

Oncor will conduct the upcoming prospective clinical trials of thebladder cancer test, and all future studies beyond Sidransky's pilottrials of other cancers.

Among these, next on his agenda are cervical cancer, because shedtumor cells are readily accessible in Pap smears, followed by lungcancer cells, recoverable in sputum. n

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.