No abstract

Background and Aims: External beam radiotherapy currently has a limited role in the treatment of hepatocellular carcinoma (HCC). The purpose of this article was to review available radiobiological data on HCC and normal liver and incorporate these data into radiobiological models that may be used to explain and improve treatment. Methods: Volume doubling times of HCC were described and used to demonstrate growth of HCC with time, assuming both exponential and logistic growth. Radiosensitivity of HCC was described and used to demonstrate the probability of uncomplicated tumor control as tumor size increases. The relationship between tolerance of liver to irradiation and volume irradiated was examined. Results: The median volume doubling time for untreated HCC was 130 days. HCC have a long period of subclinical growth. Radiosensitivity of HCC lies within the range of other tumors commonly treated with radiotherapy. When treating small volumes of normal liver, relatively high doses may be used with low risk of late radiation damage. There is a high probability of sterilizing subclinical disease and small HCC with tolerable radiation doses. Conclusion: New radiobiological data, modeling, emerging clinical data and the advantages offered by standard external beam radiotherapy techniques suggest the need for reconsidering the use of radiotherapy and for new trials.

Regular national surveys of all public and private radiation oncology facilities in Australia have been carried out between 1986 and 1999. Workforce data recorded were numbers of radiation oncologists and trainees, radiation therapists, medical physicists and physics technicians, nursing staff, data managers, social workers and clerical staff. Workloads included treatments with megavoltage beams (linear accelerators, cobalt-60), orthovoltage/superficial X-rays, brachytherapy, total body irradiation and stereotactic radiosurgery. Major equipment recorded included numbers of megavoltage and orthovoltage/superficial X-ray machines, planning simulators, computerized dosimetry systems and brachytherapy equipment. The use of radiotherapy beds and the public-private mix of treatments were also documented. Data were assembled for Australia based on each individual state. Within Australia the number of public and private treatment facilities has increased by 44% from 18 in 1986 to 26 in 1999. The population has increased by 16.4%, cancer incidence by 51.8% and megavoltage workloads (fields) by 102%. The number of radiation therapists and physicists and the number of linear accelerators have, in general, increased with the growth in workloads. The number of radiation oncologists has increased by 60% from 4.5 full-time equivalent (FTE) radiation oncologists per million population in 1986 to 7.2 per million in 1999. There is currently a deficit of at least 40 radiation oncologists to be able to treat the 50% of newly diagnosed cancer patients requiring radiotherapy. In addition, a significant deficiency exists in numbers of radiation therapists, nursing staff, data managers, social workers and clerical staff. Clearly the demands for medical physicists has increased but the data are insufficient to comment on deficiencies. Despite the increases in workloads the proportion of patients with cancer receiving radiotherapy remains below 40%. A positive correlation has been shown between the proportion of newly diagnosed cancer patients treated and the number of FTE radiation oncologists, the number of megavoltage machines and number of radiation therapists. This was shown for Australia as a whole, for each state and for the years 1986 to 1999. This was also the case when total megavoltage fields was used as the dependent variable. Multiple regression analysis using the same independent variables confirmed these positive correlations. It is concluded that the low treatment rate with radiation oncology for cancer patients in Australia is due mainly to the lack of resource allocation. The stated commitment of governments and health departments to a 50% treatment rate can only become a reality if there is a concerted effort to increase the numbers of radiation oncologists, radiation therapists, megavoltage machines and support staff. Otherwise at least one in every 10 newly diagnosed cancer patients will continue to be denied adequate and equitable access to radiotherapy - in 1999 that total figure was 9400 persons.

The demands for radiotherapy services continue to increase relentlessly but the capacity to meet these demands is not available. In Australia only 36% of patients with invasive cancer are being treated with megavoltage radiotherapy beams whereas an increase of approximately 50% is necessary to reach a more appropriate level of 55%. A review of the number of radiation oncologists, radiographers and treatment machines shows that the shortage of staff and equipment are consistent with the low utilization of radiotherapy in Australia. Staff and equipment are fully occupied and lack the capacity to appreciably increase treatment rates. A 50% increase in the number Of radiation oncologists is required to reach the recommended number of 7 per million population for service commitments only. This increase is also necessary to adequately treat the additional 50% of patients who would benefit from appropriate modern radiotherapy. Similar increases in therapy radiographer staff numbers and treatment machines will also be required if these additional patients are to be treated. A greatly expanded training, recruitment and re‐equipment programme is urgently required.

Tolerance of mature human brain to photon irradiation is described. Isoeffect curves have been derived for tolerance of large and small volumes of brain, by examination of doses determined empirically and in clinical use. These have been compared with isoeffect curves of thoracic myelitis, optic nerve and chiasm damage, and brain necrosis. The results show that the best-fitting Ellis-type equations, when five fractions per week are used, have low exponents for overall treatment time and high exponents for the number of increments (N), and are similar to published data on rat myelitis. Of the equations used to test the relationship between total dose and number of increments, the power curve was the best fit. Mean values of exponents for N derived for brain and spinal cord tolerance were 0.4 or more. These were similar to values obtained for optic nerve and chiasm damage, though the data are more limited for this complication. Brain necrosis is observed at slightly higher doses probably because of a difference in the end-point observed rather than because of any fundamental difference in tissue response. Evidence is presented to suggest that some repopulation may occur in widely spaced schedules. The use of the Ellis equation derived from connective-tissue data is inappropriate for central nervous system tissue, and its use may lead to a substantial risk of overdosage. A plea is made for adequate documentation of treatment details when human data are reported. The importance of dose per fraction is emphasized.