Research on a Novel Cancer Diagnosis and Therapy Machine



A multidisciplinary team of University physicists, engineers and radiologists headed by the University of California at Davis, and including Stanford and UCLA, is developing a source of tunable, monochromatic x-rays, suitable for the diagnosis and treatment of cancer in hospitals. The x-ray source (CXS) is based on Compton backscattering of a high peak power, short pulse laser beam off energetic electrons from a linear accelerator, producing x-rays that are tunable between 20 and 100 kV by changing the energy of the electron beam.

The CXS source is being assembled at the Stanford Linear Accelerator Center, and uses much of the linear collider technology developed there for high-energy physics research. The CXS development is part of a large “Unconventional Innovations Program” (UIP) funded by the National Cancer Institute for the last three years and aimed at high potential payoff research for the non-invasive diagnosis and treatment of cancer. Companion UIP projects to the CXS are conducting research on a variety of nanodevices (microplatforms, see for example), capable of being targeted to cancerous lesions in a patient, and being triggered to deliver drugs, or to carry contrast agents useful in radiography.

The CXS machine under development at SLAC is being designed for cancer detection and therapy, in conjunction with these UIP nanoplatforms as indicated in the schematic representation below. Tunable, monochromatic CXS x-rays are used to communicate with targeted nanoplatforms in the vicinity of cancer cells by interacting with high-Z (high atomic numbers like iodine, gold, platinum) materials attached to the nanoplatforms. The x-rays excite the K-shell of the metal, releasing low energy electrons and secondary x-rays that attack the DNA in cancer cells in the case of therapy or interact with the materials to provide a high resolution detection and monitoring capability.

The projected use of the CXS x-ray source can best be considered by the futuristic breast cancer diagnosis and treatment scenario below. The corresponding conventional scenario would involve surgical intervention in both the diagnosis and treatment phases.

A woman enters a clinic for a routine breast examination. Instead of submitting to painful breast compression, the woman lies prone on a platform, with one breast protruding through an opening in the table. A series of brief x-ray bursts enter and exit the breast from a pre-determined set of directions. Exiting x-rays are efficiently collected by a digital detector and are immediately analyzed by a computer that directly calculates breast density to a mesh volume of 30 microns on a side. By the time a radiologist reviews the images, a complete virtual 3D model of the breast has been constructed in which tissue density differences are demarcated and blood vessels and calcifications are clearly visible.

The radiologist zooms in on a region of interest, exploring the detailed morphology of the breast interior with 30-micron resolution. If warranted, an intravenous “cocktail” of “nanoplatforms”, is administered and the scan is repeated. Each nanoplatform carries a particular targeting agent and, matched to it, a high-Z metal contrast agent. Each targeting agent has specific recognition capability for the signature of a particular type and molecular classification of cancer cell. By the end of the scan, the presence or absence of a particular cancer signature is known, and the frequency of molecular recognition events can be recorded and correlated to their precise location within the breast.

Up to this point, no biopsy has been performed, but the primary site of the cancer is known, the extent of metastasis has been determined, and perhaps even the particular genetic abnormalities of the cancer have been catalogued. The diagnosis is complete, non-intrusively, and with x-ray exposures comparable to those received from a standard mammography examination.

The same CXS equipment now offers an alternative to a surgical procedure. A targeting agent is chosen for the specific type and molecular classification of cancer diagnosed, and a large dose is administered to the patient. When the molecular profile of the cancer indicates that aggressive treatment is prudent, the nanoplatform and targeting agent chosen may now also carry agents with cytoxic properties, or perhaps performing as radiosensitizers. Otherwise, less toxic compounds and concentrations may be chosen, but still with the same goal: to load the cancerous lesion with heavy metals selectively, and therefore enhance the dose deposited in the lesion when the CXS is tuned to preferentially excite the labeled agents. The same techniques as those used in conventional external beam radiation therapy for producing desired dose contours inside a patient-multiple arcs and directions of beam entry, dynamic modulation of the beam profile, and sophisticated software for treatment planning-can still be used to reduce collateral damage and non-invasively treat lesions. Efficacy is again monitored with a set of targeted contrast agents, so that alterations in the behavior of the treated cells can be detected and quantified, and further treatment adjusted appropriately. Ultimately, remission can be tracked down to the survival of small numbers of cancer cells, so that treatment neither stops prematurely, nor continues beyond what is necessary for curing the disease.

This scenario is applicable to all cancers, not just those of the breast. It is futuristic in the sense that the CXS machine has not been completed at SLAC, and the research on “nanoplatforms”, although very encouraging, has not been clinically validated. More importantly, no experiments have been performed to date using CXS in conjunction with targeted nanoplatforms. However, the principles involved are sound and have been experimentally verified by the various researchers.

It is considered that in a next phase of UIP research a completed CXS will serve as a test bed for targeting and contrast agents on nanoplatforms. Its use will greatly accelerate the development of these agents since the only alternative sources of monochromatic x-rays are research synchrotrons. These are very large multi-use machines, much less accessible than a dedicated CXS will be.

A final phase will be one in which the CXS is productized into a practical hospital clinical facility. This will require even more powerful lasers, as well as advanced microwave sources, now on the drawing board at SLAC. Such a clinical CXS, although more complex than currently used cancer therapy machines, will be undoubtedly be commercially developed if it offers the advantages described in the above scenario.




For more information, please browse the following CLS overview files.

  • Applications of the CXS to Cancer Medicine
  • Compton X-Ray Source Development



  • Use of the CLS in mammography is supported by the National Institutes of Health, Unconventional Innovations Program, and work is performed in collaboration with the University of California, Davis Medical Center.  Collaborators include Professor Dennis Matthews (LLNL and UCD DAS), Professor Johnathan P. Heritage (UCD ECE),  Dr. James E. Boggan (UCDMC Department of Neurological Surgery), and Professors Amos Norman and Timothy Solberg (UCLA).