The Four-Year Beam Delivery Experience at the M.D. Anderson Cancer Center Proton Therapy Center Houston

Reviewer: Christine Hill-Kayser
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 28, 2010

Presenter: Dr. Kazumichi Suzuki
Presenter's Affiliation: The University of Texas M. D. Anderson Cancer Center, Proton Therapy Center Houston, Houston, TX


  • The M.D. Anderson Cancer Center began delivering proton radiotherapy in May, 2006, and, since that time, > 1800 patients have been treated at their center.
    • The majority (n = 1556) have been treated with passive scattering.
    • In May, 2008, spot scanning became available, and, since that time, 260 patients have been treated with this technology.
  • The major disease sites that are treated at the M.D. Anderson Cancer Center include genitourinary malignancies (mainly prostate cancer) (52%), thoracic tumors (23%), and central nervous system tumors (22%).
  • The facility at the M.D. Anderson center operates for 12-16 hours daily, 6 days per week. Approximately 100 patients are treated with proton therapy there daily.
  • Currently, in the United States and worldwide, particle therapy is a limited medical resource, with many more patients having indications for this type of treatment than can receive it.
  • In addition, as economic resources continue to shrink, improving healthcare efficiency is of the utmost importance.
  • In order to address these needs, the authors carried out the study presented here in order to assess patient census and uptime of the facility, and to identify ways in which efficiency might be improved in order to improve machine efficiency and, thus, increase availability to patients.


  • All beam delivery parameters for every treatment field delivered have been maintained in a database at M.D. Anderson. These include:
    • Delivered dose
    • Energy
    • Range
    • Spread out Bragg peak width
    • Gantry angle
    • Couch position
  • In addition, delivery system downtime for every equipment failure has been recorded and maintained.
  • These data were used to evaluate the frequency of beam delivery parameters, patient census, and uptime of the facility operation.
  • A subset of 75 patients was identified for further analysis of duration of each treatment session from patient walk-in to walk-out.
    • Patients had genitourinary, thoracic, and central nervous system malignancies
    • They were treated with variable numbers of fields, and were of varied age and performance status.


  • The yearly averaged facility uptime in 2009 was 96%, which exceeded the target value of 95%.
  • Analysis of the comprehensive database showed that approximately 6000 fields have been delivered.
  • The average yearly delivered dose was 10 kilogray (kGy) for each of the four treatment rooms.
  • The subset of 75 patients was then analyzed.
    • The facility process was identified to be as follows:
      • Patient enters facility
      • Patient undergoes anesthesia if indicated
      • Patient is immobilized
      • Set-up images are obtained
      • Equipment is set-up
      • Devices are loaded/unloaded
      • Beam is requested
      • Treatment is delivered
      • Patient is unloaded and exits
    • The above process was divided into patient oriented activities (patient entering, undergoing anesthesia, being immobilized, being set-up, being unloaded), equipment activities (set-up of equipment and loading/unloading of devices), and beam delivery (requesting of beam and delivery of treatment).
    • On analysis of these subsets of process time for these 75 patients, 80% of comprehensive treatment time was found to be dedicated to patient and equipment activities (43% patient oriented activities, 38% equipment activities). The remaining 19% of time was dedicated to beam delivery.
    • The authors examined the total treatment time as a function of the number of fields for these 75 patients.
      • They found that treatment time increased with the number of fields via a quadratic relationship.
      • Variation among patients treated with the same number of fields as one another depended on patient performance status.

Authors' Conclusions

  • The authors conclude that their facility has operated at a high performance level, treating a large number of patients with a variety of disease sites.
  • They note that the results of this study have been useful for their operation planning, and call on vendors and collaborators to work to increase efficiency related to patient and equipment factors.

Clinical/Scientific Implications

  • The authors have performed an interesting study examining the efficiency of their center, as well as performing an in-depth examination of the ways in which time is distributed during patient treatments.
  • Their finding that the majority of time spent treating patients is related to patient and equipment factors is not altogether surprising. Their point that future work should focus on reduction of time delays in patient immobilization, set-up, and equipment exchange is well taken. Certainly, tools such as multileaf collimators that prevent need for placement of heavy blocks may contribute to these goals.
  • Their findings regarding influence of patient performance status on overall treatment time are also of interest, as findings such as these may contribute to decisions about distribution of appointment times throughout the day, as well as staffing issues for patients who may need more assistance due to lower performance status.
  • In the economic milieu in which we currently exist, attention to efficiency in the medical world is of utmost importance. Certainly, our collective goal should be to increase accessibility of this limited resource to all patients who have need of it. Preventing unnecessary time delays will certainly contribute to this.