Secondary Cancers After Radiotherapy: An Increasing or Decreasing Concern?

Eric Shinohara, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: August 8, 2007

Presenter: Eric J. Hall
Presenter's Affiliation: Columbia University
Type of Session: Scientific


  • Approximately 10% of patients presenting to US cancer centers have secondary malignancies.
    • This may be due to an increase genetic susceptibility to cancer or continued life style habits which put patients at increased risk for second malignancies.
    • Iatrogenic second malignancies are of great concern.
  • Secondary malignancies are of greater concern due to the increasing number of long term survivors who have been treated with radiation. It is generally difficult to evaluate the risk of secondary malignancies within the context of the baseline incidence of cancer. However, in cervical and prostate cancer, where both surgery and radiation are common treatment options, there is an appropriate control group.
    • The ten year survival of breast and prostate cancer patients is approximately 76%
    • Dr. Hall has previously published that the estimated risk of developing a second malignancy in patients treated with radiation for prostate cancer is approximately 1/70 for those who survive beyond 10 years based on SEER data.
      • This study also found that up to 1/3 of secondary cancers occurred far from the treatment field.
    • Studies in patients treated with radiation versus surgery for cervical cancers have also shown an increased risk of secondary malignancies in those patients treated with radiation.
  • Theoretically, highly conformal radiation therapy, such as IMRT, should decrease the dose to normal tissue and allow increased dose to the tumor. This should reduce the risk of secondary malignancy due to the reduced dose to normal tissues.
    • However, IMRT requires more fields and more monitor units (2-3 fold more) to deliver treatments.
    • This increases the total body dose due to radiation leakage from the head and multi-leaf collimators (MLC) of the linac.
      • Leakage from the head of the linac is approximately 0.1% of the dose at isocenter.
      • Leakage from the MLC's is estimated to be 1.5 to 3.0% of the dose at isocenter.
    • Studies using a phantom were performed which measured the dose the trunk received when 70 Gy was given in the region of the prostate/cervix.
      • For a 6 MV beam, the dose from photons to the trunk was found to be approximately 25 cGy.
      • For a 20 MV beam the dose from photons to the trunk was found to be approximately 50 cGy, with an additional 1 cGy in neutron dose. This was equivalent to an exposure of 0.75 Sv.
      • A dose of 0.75 Sv is within the range of exposure of atomic bomb survivors (approximately 0.05 to 2.5 Sv).
  • How the dose of radiation affects the risk of secondary cancers has been studied previously. Initial studies by Gray suggested that the risk of secondary cancers increased gradually with dose but decreased at higher doses due to cell death.
  • However, how do larger, fractionated doses affect the risk of secondary malignancy? Dr. Hall and others have found that at higher doses the risk of secondary cancer does not appear to decrease but rather reaches a plateau.
  • Dr. Hall has previously shown that IMRT appears to double the relative risk of secondary cancer compared with 3D conformal treatment, from 1.5 to 3.0. This is in agreement with data recently published by MD Anderson.
    • This increase is thought to be due to an increase in monitor units and an increase in the volume of normal tissue treated.
    • The increased risk was found to have two peaks, one at low doses and one at very high doses. Protons may eliminate the low dose peak.
    • The 2 fold increase in relative risk for secondary malignancies may be acceptable in older patients, such as prostate cancer patients, if acute toxicity and local control can be improved.
    • However, the relative risk of secondary malignancies is higher in children.
      • Children have an approximately ten fold higher sensitivity to radiation. Prior studies have shown that exposures on 0.5 to 1 Sv leads to a 7-15% increase in the risk of developing secondary cancers in children treated with radiation.
      • Children are smaller with organs that are more compact which may increase organ exposure to radiation.
      • Children are more likely to have genetic defects which may make them more susceptible to developing secondary malignancies.
  • IMRT is a useful technique which improves the conformality of dose around tumors and its use should be continued.
    • However, the linacs currently in use were not designed for IMRT and the amount of shielding in the head needs to be increased to reduce scatter dose to the patient.
    • Secondary beam blocking for the scatter from MLC's needs to be implemented to further limit scatter dose.
    • Flattening filters are not necessary for IMRT and should not be used. This will decrease scatter and increase the dose rate.
  • Protons are the next step in improving conformality and should further decrease dose to normal tissues. However, there are issues with total body dose from neutrons.
    • To create a beam of an adequate size for clinical use scatter foils are commonly used. When protons interact with the scatter foil neutrons are created which increases the total body dose of radiation to the patient.
  • Dr. Hall addressed the debate regarding his recently published paper (Int. J. Radiation Oncology Biol. Phys., Vol. 65, No. 1, pp. 1–7, 2006) examining dose outside the field's edge from IMRT, proton treatment using scatter foils, and proton treatments using a scanning beam.
    • The dose outside the field's edge approximately doubles in patients who receive IMRT compared with 3D conformal therapy.
    • The dose outside the field's edge from a scanning proton beam is less than with IMRT.
    • The data for the proton beam utilizing scatter foils obtained from Harvard was old. The true value for the dose outside of the field's edge for proton beams utilizing a scatter foil may be 9 fold less than reported in Dr. Hall's paper. This is close to the doses seen in IMRT treatment.
    • However, Dr. Hall believes that the relative biological effect (RBE) of neutrons used in the estimate of dose was too low at 10 and believes that an RBE of 30 would be more appropriate. This puts the dose outside the fields edge of proton beams which use scatter foils above that of IMRT, however it is still less than the dose Dr. Hall initially published.
    • Studies have demonstrated that the bulk of neutrons created in proton therapy are from the scatter foils and hence the use of a scanning beam, which does not require the use of scatter foils, should significantly reduce neutron production.
    • It is possible that the benefit from the increased conformality that protons provide may be reduced by the increase in dose to normal tissues from neutrons. Therefore, when possible scanning beams should be used.

Author's Conclusions

  • The current standards for head shielding were defined when predominantly 3D conformal therapy was used. New, stricter shielding parameters are needed as IMRT uses more fields and monitor units, leading to an increase in total body dose from scatter.
  • When using protons, scanning beams should be used when possible to reduce total body dose from neutrons.

Clinical/Scientific Implications

  • With further improvements in cancer detection and treatment, patients will be diagnosed earlier and survive longer. With these improvements, the concern shifts to the long term effects of radiation, particularly secondary malignancies.
  • The use of IMRT has expanded greatly over the last 5 years. Due to the improvements in conformality, it is heavily used in the pediatric population, which has an increased susceptibility to secondary cancers. There is an increasing body of evidence suggesting that total body dose is increased and that stricter shielding parameters are needed.
  • Proton use will also increase as the numbers of proton facilities continues to expand. Scanning beams should be used when possible to reduce neutron dose. However, this is complicated by organ motion. Currently the ability to treat moving tumors with scanning beam protons is limited. In the future, gating and other techniques may allow the use of scanning proton beams in mobile organ tumors, however, currently treatment with proton beams which utilize a scatter foil are needed for these types of tumors.
  • Further studies measuring the actual neutron dose from modern proton therapy units should continue to be performed for both scattered and scanned beams.

Frequently Asked Questions