Oak Ridge National Laboratory


News Release

Media Contact: Carolyn Krause ()
Communications and External Relations


Cancer patients may benefit from ORNL computer code

OAK RIDGE, Tenn., April 10, 1996 — Patients needing bone marrow transplants or radiation therapy to kill cancerous cells may benefit from a computer code developed by the Department of Energy's (DOE) Oak Ridge National Laboratory (ORNL).

The code, called MarCell for marrow cells, helps physicians plan a safe course of treatment for patients needing bone marrow transplants to replace their immune systems or radiation therapy to combat cancer. Using the code, originally developed for the Department of Defense (DOD) and the North Atlantic Treaty Organization (NATO), researchers can calculate the probability that a patient will develop cancer or die within 30 days of radiation treatment of bone marrow. DOD and NATO were interested in the code to predict the survival rate for soldiers exposed to radiation for weeks or months during nuclear war.

MarCell has also sparked interest by NASA's Langley Research Center in Hampton, Va. The center will be using MarCell to determine the risk to astronauts who are exposed to radiation from solar flares and cosmic rays in outer space. Astronauts are exposed to space radiation during shuttle flights and during work on orbiting space laboratories. NASA plans to use MarCell to determine shielding needs and formulate guidelines for altering or ending an astronaut's space activities.

MarCell was developed by Troyce Jones, a researcher in ORNL's Health Sciences Research Division; Max Morris, a research statistician in the Computer Science and Mathematics Division; and Jafar Hasan, who worked as an undergraduate with Jones. Hasan, whose work was sponsored by DOE and ORNL under the Great Lakes Colleges Association/Associated Colleges of the Midwest Oak Ridge Science Semester, recently graduated from Albion College in Michigan. He has continued to work on the code as a consultant to ORNL.

"The development of the code led to a revolutionary finding about the nature of bone marrow," Jones said. "We found that the cells that were most radiosensitive, the stem cells, seem to be considerably less important than previously believed. We have found that the cells that are the most important - in fact, critical to the production of blood cells - are cells of the marrow stroma, including fibroblasts.

"Fibroblasts and even stromal cells have traditionally been considered the least important cells in the complex process of blood formation. Our calculations support recent experimental evidence that stromal cells are critical to this process."

"We think our code also can be helpful for doctors and patients planning bone marrow transplants," Jones said. "It can predict the risk of total body cancer and leukemia from various sources, levels and dose rates of radiation from medical, occupational and environmental exposures."

Recently, researchers have used data from patients with leukemia and lymphoma to determine the response to radiation treatment of malignant cells compared with normal stem and stromal cells. For malignant cells, cell proliferation rates or cancer doubling time data can be entered for the individual patient.

Information gained through these studies should provide physicians with better tools for treating patients with leukemia or aplastic anemia, a disease that prevents bone marrow cells from producing blood cells.

"The user-friendly code helps you decide how many radiation treatments are needed, how to minimize risk to the patient and how long the recovery period will last for a particular course of treatment," Jones said. "It provides some information on patient responses to different therapeutic aids, such as antibiotics and blood transfusions."

MarCell models the rate of cell loss and recovery for marrow stromal cells and stem cells exposed to 12 types of radiation, including X-rays, gamma rays, neutrons, beta and mixed fields of neutron-gamma rays.

To use the code, the operator simply enters dose, dose rate, number of exposures and time between each exposure. A computer-generated graph then shows the number and rate at which bone marrow cells are dying. An option permits the same graph to show how different cell lineages repopulate during, between and after individual radiation treatments; some of the injured marrow cells may proliferate and replace some of the normal bone marrow stem cells. MarCell calculates the increased risk of cancer of the blood and lymph glands that results from marrow transfusion or immunosuppression.

ORNL researchers were the first to use modern statistical techniques and sophisticated computing power to address the thorny problem of the death rate and growth rate of irradiated bone marrow cells.

"Our big surprise was that the results of Morris' calculations did not describe stem cells as the cells most critical to the survival or death of an animal," Jones said. "I suggested that stroma cells might be more critical than believed, and the results on both animal and cell survival rates matched those of cells in the mathematical model."

This finding has been confirmed by more recent experimental work, according to Jones, who said the evidence shows that stroma cells (stroma means "bed" in Greek) do much more than simply serve as a supportive layer to which stem cells must attach before they can proliferate.

"These fatty yellow cells produce growth factors, or cytokines, that tell stem cells when and how fast to divide and how to differentiate into platelets and red, white and other kinds of blood and lymphatic cells," Jones said.

The ORNL researchers' code calculations have led to 10 papers in a number of journals, including the International Journal on Radiation Oncology.

ORNL's work in this area has generated interest in applying the same methods to model losses and recovery of irradiated lymphocyte cells. Such information can serve as an early biological indicator of the response of victims of radiation accidents - such as those exposed to radiation from the Chernobyl reactor accident - as well as the response of patients to a series of therapeutic treatments.

The code runs on a DOS- or Windows 95-based operating system and works effectively on a Model 286 or better.

ORNL, one of DOE's multiprogram research laboratories, is managed by Lockheed Martin Energy Research Corporation.