The Value of Dose Constraints for Organs at Risk in Proton Therapy for Prostate Cancer

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

Presenting Author/Institution: Nancy Mendenhall, MD, University of Florida Proton Therapy Institute


  • Prostate cancer is the leading cancer diagnosis in men. Treatment options include radical prostatectomy, radiotherapy, and/or hormone therapy.
  • Proton radiotherapy for prostate cancer has the potential to spare dose the bowel and bladder due to is stopping power at the distal edge of the beam.
  • For intensity-modulated radiation therapy (IMRT), dose-volume parameters have been found to be predictive of bowel and bladder toxicity.
    • A recent study by Pedersen et al (IJROBP, 2012) showed that all men were free from Grade 2+ GI toxicity at 4 years if rectal V(70) ? 10%, V(65) ? 20%, and V(40) ? 40%.
  • For proton beam therapy, the dose-volume parameters needed to reduce toxicity are unknown.
  • The goal of this study was to assess the usefulness of dose-volume constrains for minimizing the risk of toxicity with image-guided proton therapy for early and intermediate risk prostate cancer


  • This study analyzed 171 prostate cancer patients who were treated from 2006-2008 on two prospective IRB-approved trials:
    • Low-risk protocol delivering 78 Gy (RBE) in 39 fractions, 2 Gy (RBE) per fraction.
    • Intermediate-risk dose escalation protocol delivering 78 to 82 Gy (RBE) in 2 Gy (RBE) fractions.
      • Of these patients, 70 received escalation to 80 Gy (RBE) (13) or 82 Gy (RBE)(52)
  • Minimum potential follow up was 3 years. Median follow up was 4 years.
  • Patients were immobilized with placement of Visicoil markers for daily image guidance.
    • Bladder filling and rectal saline was used and rectal balloons were only used when prostate motion exceeded PTV motion on an as-needed basis.
  • Target structures and organs-at-risk (OARs) were delineated on CT with aid of MR fusion. MR was found to be particularly useful for prostate apex. Gross tumor volume (GTV) was defined as prostate only in the low-risk protocol and was prostate and seminal vesicles in the intermediate- risk protocol.
  • GTV was expanded to create the planning tumor volume (PTV) by 8 mm in the superior-inferior axis and 5 mm axially.
  • Individualized beam angles were selected to minimized variable skin surface contour. The following beam placements were used: 6% opposed laterals, 42% anterior obliques, 45% posterior obliques, and 7% mixed.
  • Aperture margins were 1 cm superior, inferior, and anterior and 7 mm posteriorly.
  • The distal/proximal margins were 0.5 cm from PTV. Distal compensator smearing was 1.9 cm.
  • The following dose-volume requirements were set:
    • 95% of PTV receives 100% of the dose and 100% of the PTV received 95% of the dose
    • The rectum (RE) and rectal wall (RW) had separate DVH parameters. It was required that the RE V70 be less than 30% and the V50 be less than 50%
  • Toxicity rates were correlated with clinical factors and prospectively calculated dose volume relationships in organs at risk.


  • Regarding cancer control, 1 low risk and 1 intermediate risk patient had disease recurrence. Therefore, PFS/BFS was 99% at 3 y. Specifically, the recurrences were two isolated pelvic nodal recurrences.
  • Overall survival was 95% at 3 y. There were 14 deaths from intercurrent disease (DID), 8 on low-risk protocol and, 6 on the intermediate-risk protocol.
  • The prevalence and cumulative incidence of grade 2+ GI toxicity at 3 years were 5.8% and 12.9%, respectively. Actuarial 3 year projection of grade 2+ GI toxicity was 13.1%.
  • GI toxicity was manifested by rectal bleeding, proctitis, rectal incontinence, diarrhea, and abdominal pain.
  • There were two grade 3 GI toxicities which included one after rectal biopsy leading to a colostomy and the other a blood transfusion.
  • Univariate analysis was performed, examining the effect of age, protocol, comorbidities, medications, and DVH parameters including rectal wall (RW) and rectum (RE) on the incidence of grade 2+ GI toxicity.
    • Only RE V30-V75 had statistically significant correlations.
  • Multivariate analysis showed significant correlation between grade 2+ rectal bleeding and proctitis and V30-V75: V 30(p=0.04), V70 (0.013) and V75 (0.012).
  • There were statistically significant differences in the incidence of GI toxicity in rectal V30 <38% vs > 38%, V70 <21% vs > 21%, and V75 <15% vs > 15%.
    • The authors identified that the V75 constraint was the most statistically significant.

Author's Conclusions

  • Outcomes at 3 years with image-guided proton therapy in doses up to 82 Gy (RBE), based on organ-constraints and the techniques used in this study, continue to suggest high efficacy and minimal toxicity.
  • The correlation noted between grade 2+ rectal bleeding and/proctitis and various dose-volume parameters may be useful in designing dose-volume constraint goals for organs at risk in future clinical trials.
  • The most powerful dose-volume predictor of grade 2+ GI toxicity is V75, which should be maintained <10-15%. The impact of volume received by low-intermediate doses is controversial; to be prudent, V30 should be less than 40% and V70 less than 20%.

Scientific/Clinical Implications

  • This study presents one of the first long-term follow up studies assessing the risk of GI toxicity and DVH parameters with proton beam therapy. They show excellent cancer control outcomes and low rates of toxicity.
  • The DVH parameter that most strongly correlated with GI toxicity is the V75, which is consistent with other studies examining GI toxicity and IMRT. It is of note that this constraint was not empiric at the time of planning. Statistically, it is very challenging to sort out if the V75 alone is responsible for toxicity, or if other low and intermediate doses such as V50 or V30 are also critical as the data is nested within the V75 data. Therefore, we agree with the authors' conclusions that V75, V70, and V30 be minimized in proton treatment planning.
  • A main strength of this study is that varied gantry angles were used to reduce the topographic variability of the patient.
  • The rectum was defined in this study from sigmoid flexure to the bottom of the ischial tuberosities. Many institutions use a more stringent definition of the rectum which would make meeting these constraints more challenging.
  • There are additional strategies that can be employed to decrease PTV margins and therefore spare the rectum further. They include:
    • Improved patient and prostate immobilization techniques, e.g. rectal balloon to permit smaller PTVs
      • The University of Florida and the University of Pennsylvania now uniformly use rectal balloons with a smaller PTV margin than utilized in this study.
    • Adaptive therapy with weekly CBCT to monitor for changes in bladder or bowel size.
    • Pencil beam scanning to improve the conformality of proton beam therapy.
  • Future directions could include normal tissue complication probability (NTCP) modeling as well as a randomized trial with the same target coverage and varied constraints on the bowel.


I wish u knew... Understanding prostate cancer screening
by Timothy J. Hampshire
December 06, 2012

Cancer Center Advertising: Doing More Harm Than Good?
by Rodney Warner, JD
July 21, 2016