Tumor Stem Cells May Improve Research on Brain Tumors

Researchers at NCI and the National Institute of Neurological Disorders and Stroke have found that glioblastoma tumor stem cells are a better model for studying the biology and physiology of glioblastomas than the cancer cell lines typically used in research laboratories. Tumor stem cells accurately reflect the biological mechanisms and genetic characteristics of the parent tumor, the researchers report in the May 15 Cancer Cell.

"This study illustrates that traditional cancer cell lines are a flawed model and poorly represent human tumors," said lead researcher Dr. Howard Fine, chief of NCI's Neuro-Oncology Branch. "We have shown that these tumor stem cell lines may ultimately offer a model system that more accurately represents the biology of the tumors actually found in patients."

His team grew glioblastoma tumor stem cell lines in two different culture conditions: one containing serum and the other serum free. The resulting stem cell lines were then compared with traditional glioblastoma cell lines grown in serum and with normal neural stem cell lines grown in serum-free solutions.

The investigators found that the glioblastoma stem cells grown under the two different conditions had very different physiological and genetic characteristics. The tumor stem cells grown under serum-free conditions had the same characteristics as the parent glioblastoma cells. The glioblastoma tumor stem cells cultured with serum, however, lost all biological and genetic characteristics of the original tumor cells.

Computerized Ordering of Chemo Drugs Slashes Error Risk

Use of a computerized system to order chemotherapy for pediatric cancer patients dramatically reduces the risk of potentially dangerous medication errors, according to a new study from Johns Hopkins University researchers.

To conduct the study, published in the May issue of Archives of Pediatric & Adolescent Medicine, the researchers conducted daily audits of sequential pediatric chemotherapy orders before and after the deployment of a new computerized provider order entry (CPOE) system in the Children's Center at Hopkins' Sidney Kimmel Comprehensive Cancer Center.

Compared with handwritten orders, use of the CPOE system resulted in a 74-percent reduction in improper dosing, a 91-percent reduction in incorrect dosing calculations, and a 68-percent reduction in missing cumulative dose calculations.

Multiple checks on medication orders by oncologists, pharmacists, and nurses are already in place, says lead author Dr. George R. Kim, so the chances of the errors getting through to patients are small.

Nevertheless, says co-author Dr. Allen Chen, dosing errors are potentially more harmful in children because, among other things, they absorb and metabolize drugs differently. "It's something that we worry a lot about in pediatrics because the dose range that's valid for one patient or another is enormous," he says.

Implementation of the CPOE system, the researchers cautioned, also introduced a new error: Medications were ordered that were not included in patients' approved treatment plans. The finding, Dr. Chen notes, reinforces the "importance of measuring the results of change and the need for careful monitoring when introducing a new system."

The results from the Hopkins trial, says Dr. Crystal Mackall, acting chief of the NCI Pediatric Oncology Branch, demonstrate that migrating clinical care processes - such as ordering medications - to a computer-based system with built-in decision support "should be a matter of high priority for improving the delivery of cancer care."

Rise in Thyroid Cancer Attributed to Better Detection

Data from NCI's Surveillance, Epidemiology, and End Results (SEER) program suggest that thyroid cancer in the United States increased by a factor of 2.4 between 1973 and 2002, rising from 3.6 to 8.7 cases per 100,000. Writing in the May 10 Journal of the American Medical Association, Drs. Louise Davis and H. Gilbert Welch of the Veterans Affairs Outcomes Group in White River Junction, Vt., conclude that "increased diagnostic scrutiny has caused an apparent increase in incidence of cancer rather than a real increase."

The authors found that papillary cancer (which accounts for 88 percent of all thyroid cancer) increased 2.9-fold, from 2.7 to 7.7 cases per 100,000, accounting for virtually all of the observed increase in overall incidence between 1973 and 2002. The increase was almost entirely attributable to small papillary cancers: Since 1988, 87 percent of the increase came from cancers 2 cm or smaller, and 49 percent from papillary thyroid microcarcinomas 1 cm or smaller. Improvements in technology made these smaller cancers increasingly likely to be detected; thyroid ultrasound can detect nodules as small as 0.2 cm, and fine-needle aspiration can immediately assess the cytology of the tissue. Since mortality remained constant throughout the 30-year period, the most likely conclusion is that these nodules were there all along, unobserved, underdiagnosed, and not progressing to death or even clinical disease.

In an accompanying editorial, Dr. Ernest L. Mazzaferri of the University of Florida agreed with the analysis, but warned that "some microcarcinomas are destined to become aggressive cancers," and cannot be ignored or relegated to watchful waiting on the basis of small size alone.

Researchers Identify Prostate Cancer Gene Variant

Researchers are reporting what they say is the first identification of a common genetic variant directly linked to the increased risk of prostate cancer. The variant, they reported early online May 7 in Nature Genetics, is more prevalent in African American men than in those of European ancestry, which, they note, may explain why African Americans are at higher risk of the disease.

Led by Dr. Kari Stefansson, of deCODE Genetics in Reykjavik, Iceland, the study included genetic analyses of participants in four case-control study groups from Iceland, Sweden, and the United States. They identified a genetic marker on chromosome 8, an allele of a microsatellite DG8S737, that increased risk in these different groups by more than 60 percent. The marker's presence yielded a "population attributable risk" - that is, the expected percentage of a given population whose cancer is attributed to the variant - of 8 percent in those of European ancestry and 16 percent in African Americans.

In the study, the researchers initially identified a suspect section on chromosome 8 in a study of Icelandic men with prostate cancer. They then performed genotyping on samples from participants in the four prostate cancer case-control studies to further determine if they could pinpoint a genetic variant linked to risk.

Compared with controls, the variant's frequency was significantly greater in prostate cancer patients with Gleason scores of 7 to 10. The variant's frequency also was greater in prostate cancer patients with higher Gleason scores than in those with lower Gleason scores, but only to a modest degree.