How much is the Proton Range Uncertainty? An End-to-end Study Using Various Animal Tissues

Reporter: Abigail Berman Milby, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 21, 2012

Presenting Author/Institution: Yuanshui Zheng, ProCure Proton Therapy Center


  • The physical advantage of proton therapy is that particles have a finite range and, therefore, normal, healthy structures that are beyond the area of tumor can be spared from unnecessary radiation.
  • Proton beam therapy is increasingly being adapted as a mechanism to decrease damage to normal tissue structures and the risk of secondary malignancies. Proton beam therapy is used in a wide range of disease sites, including prostate, gastrointestinal, lung, head and neck, and central nervous system malignancies.
  • Due to the favorable decreased integral dose and normal tissue sparing, proton therapy is widely used in the treatment of pediatric malignancies.
  • One caveat to proton beam therapy is that there is range uncertainty which can lead to target underdose or normal tissue overdose. Accurate assessment of range uncertainty is critical in proton beam therapy.
  • Proton therapy practitioners address range uncertainty by avoiding directly aiming a single beam at critical structures and adding margins in the range uncertainty direction.
  • Margins can be calculated by incorporating the range uncertainty from the CT, which is the uncertainty in the CT number and its conversion to stopping power, and the range uncertainty due to tissue and patient specific variations.
  • There is a lack of data and consensus on what the appropriate amount of uncertainty is. Range uncertainties used at proton therapy centers range from 1-3.5%.
  • This study quantified the range uncertainty differences between the treatment planning system (TPS) and actual delivered dose through various animal tissues.


  • Animal tissues including a pig head, beef steak, and lamb leg were used in this study. For each tissue, an end-to-end test closely imitating patient treatments was performed.
  • CT simulation, treatment planning, image-guided alignment, and beam delivery was performed for each animal tissue.
    • CT scanner used was Lightspeed RT GE (Wisconsin) with 1.25 mm slice thickness.
    • Treatment plans were created using XiO TPS (CMS, St. Louis, MO).
    • Two beams were used in each plan, one with a tissue specific compensator and the other without it.
    • An orthogonal x-ray imaging system was used to align the tissue before a proton beam of 200 monitor units was delivered.
  • Five pieces of Gafchromic EBT2 films were placed at various depths in the distal dose falloff region.
    • The film was calibrated using ionization chamber measurements.
  • Comparison between measured and calibrated doses was performed, and used to calculate range differences.


  • The dose difference at the distal falloff between measurement and calculation depended on tissue type and treatment conditions.
  • The maximum dose differences were as follows:
    • In pig head irradiation, there was a difference of 35% which corresponds to a range difference of 3.5 mm
    • In beef steak irradiation, there was a difference of 54% which corresponds to a range difference of 5.5 mm.
  • The range differences were within 2% for all available tests.
    • The TPS was able to calculate the proton range within 2%, or 1.5% plus a 1.5 mm margin for non-moving targets.

Author's Conclusions

  • The TPS was able to calculate the proton range within 2%, or 1.5% plus a 1.5 mm margin for non-moving targets. This is smaller than the standard uncertainty of 3.5% plus 3 mm uncertainty used at most operating proton centers.
  • Further studies need to be performed to detail the causes of the range uncertainty and the relative contributions to the uncertainty.
  • Accurate assessment of range uncertainty in treatment planning would allow better optimization of proton beam treatment, thus fully achieving proton beams' superior dose advantage over photon therapy.

Scientific/Clinical Implications

  • This study supports a smaller distal range uncertainty than is standardly incorporated into proton treatment planning worldwide. This data is robust, using multiple types of animal models and serial measurements at the distal uncertainty region.
  • Further studies will need to be performed to reproduce this data to universally change the practice of 3% range uncertainty plus 3 mm margin to 1.5% plus 1.5 mm margin.
  • Clinically, this represents an important advance in proton therapy treatment planning. Often, the tumor target structure is directly adjacent to a critical organ, such as the bowel, spinal cord, or optic nerve. With the current state of range uncertainty in proton treatment planning, adjacent normal structures often will have to receive the full prescribed dose because of the margin incorporated for uncertainty. The fall-off of the proton beam is often within, or just beyond, a critical structure. This limits the total dose that practitioners can give to tumors directly adjacent to critical structures as we are limited by the maximum dose tolerance of that structure.
    • If range uncertainty at the distal beam edge can be reduced, as suggested in this study, dose escalation of tumors adjacent to critical structures may be possible.
  • Future studies will need to address how the range uncertainty varies with 1) physical location of tumor; 2) biology of tumor and normal tissues; 3) proton unit; 4) TPS and CT scan.
    • This study addresses the physical range uncertainties only. There is evidence as well to suggest that relative biologic effectiveness (RBE) varies at the distal edge of the proton beam as well, adding a separate uncertainty that is not address in this study.
    • Further studies will also need to elucidate the differences in range uncertainty with passive scatter proton therapy versus pencil beam scanning and intensity-modulated radiation therapy.