J Roll

Radiotherapy quality assurance of gynecologic oncology group (GOG) protocol 165, a cooperative group study of carcinoma of the cervix

ABSTRACT: GOG protocol 165 was a randomized phase III trial to evaluate radiation vs. radiation plus weekly cisplatin vs. radiation plus protracted venous infusion 5-FU in patients with stage II-B, III-B and IV-A carcinoma of the cervix. Protocol treatment included external beam therapy and high-dose rate or low-dose rate brachytherapy. Historically, GOG has performed an extensive review of patient treatments on their clinical trials. As HDR had no previously been used in GOG trials, credentialing of institutions and physicians was required prior to entering patients onto the study for the use of HDR, but was not required for external radiation therapy or LDR. Credentialing consisted of a review of the institution’s HDR physics and QA, and a clinical and dosimetric review of the brachytherapy treatment of two patients treated by the same radiation oncologist in a manner similar to the protocol guidelines. The credentialing process not only evaluated the quality of HDR procedures at the institution, but also assured that the institution and participating radiation oncologist had HDR experience. At the same time, it educated the institution as to the specific requirements of the protocol. Retrospective review of radiotherapy of 326 patients entered on the study was performed. A recalculation of patient dose and a review of the records and all planning and verification films were performed by the Radiological Physics Center in conjunction with the GOG HDR subcommittee and the protocol study co-chairs, respectively. Deviations from protocol guidelines were assessed according to predefined criteria. 100% of the patients treated at credentialed institutions were treated without major protocol deviations. In contrast, 81% of patients from non-credentialed institutions completed treatment without major deviations. Minor deviations occurred in both groups with the result that 75% and 50% of patients from credentialed and non-credentialed intuitions respectively were treated in strict compliance with the protocol. A breakdown of protocol deviations appears in table 1. Table 1. Summation of Deviations for Non-Certified and Certified Institutions

WE‐C‐T‐617‐09: The Radiological Physics Center's Anthropomorphic Quality Assurance Phantom Family: New Developments

Purpose: To develop quality assurance (QA) phantoms that simulate specific treatment sites for the purpose of credentialing institutions for participation in NCI‐sponsored advanced technology clinical trials. Method and Materials: The Radiological Physics Center (RPC) has developed an extensive credentialing program for institutions wishing to participate in clinical trials that use advanced technologies such as IMRT. This program includes questionnaires and irradiation of an anthropomorphic QA phantom specifically designed for the trial. These QA phantoms typically are water‐filled plastic shells with imageable targets, avoidance structures, and heterogeneities that contain TLD and radiochromic filmdosimeters. Three new trials are presently under development; 1) extracranial stereotactic radiotherapy for liver metastases, 2) IMRT for endometrial or cervical cancer and 3) advanced RT for mesothelioma. New phantoms or target/dosimetry inserts have been designed to meet the credentialing needs for the three trials listed above. Results: A new liver phantom has been designed and constructed. The liver phantom contains two targets within the liver and three organs at risk (OAR); the stomach, kidney and spine. This phantom will be placed on a 2D reciprocating table to include target motion in the AP and SI directions. A new target/dosimetry insert has been designed for the RPCs existing thorax phantom that will include an OAR for the liver and dosimeters in the lung and in the chest wall. Finally, another target/dosimetry insert has been designed for the pelvic phantom that will include a vagina, cervix, bladder and rectum. Conclusion: The RPC has been and will continue to be proactive and responsive to the needs of the study groups as new treatment modalities are used for new clinical trials. This work was supported by PHS grant CA10953 and CA081647 awarded by NCI, DHHS.

SU-E-T-148: Benchmarks and Pre-Treatment Reviews: A Study of Quality Assurance Effectiveness

Purpose: To determine the impact benchmarks and pre-treatment reviews have on improving the quality of submitted clinical trial data. Methods: Benchmarks are used to evaluate a site’s ability to develop a treatment that meets a specific protocol’s treatment guidelines prior to placing their first patient on the protocol. A pre-treatment review is an actual patient placed on the protocol in which the dosimetry and contour volumes are evaluated to be per protocol guidelines prior to allowing the beginning of the treatment. A key component of these QA mechanisms is that sites are provided timely feedback to educate them on how to plan per the protocol and prevent protocol deviations on patients accrued to a protocol. For both benchmarks and pre-treatment reviews a dose volume analysis (DVA) was performed using MIM softwareTM. For pre-treatment reviews a volume contour evaluation was also performed. Results: IROC Houston performed a QA effectiveness analysis of a protocol which required both benchmarks and pre-treatment reviews. In 70 percent of the patient cases submitted, the benchmark played an effective role in assuring that the pre-treatment review of the cases met protocol requirements. The 35 percent of sites failing the benchmark subsequently modified there planning technique to pass the benchmark before being allowed to submit a patient for pre-treatment review. However, in 30 percent of the submitted cases the pre-treatment review failed where the majority (71 percent) failed the DVA. 20 percent of sites submitting patients failed to correct their dose volume discrepancies indicated by the benchmark case. Conclusion: Benchmark cases and pre-treatment reviews can be an effective QA tool to educate sites on protocol guidelines and to minimize deviations. Without the benchmark cases it is possible that 65 percent of the cases undergoing a pre-treatment review would have failed to meet the protocols requirements.Support: U24-CA-180803

TU‐G‐BRD‐06: The Imaging and Radiation Oncology Core Houston (IROC Houston) QA Center International Activities Outside North America

Purpose: To describe the extent of IROC Houston’s (formerly the RPC) QA activities and audit results for radiotherapy institutions outside of North America (NA). Methods: The IROC Houston’s QA program components were designed to audit the radiation dose calculation chain from the NIST traceable reference beam calibration, to inclusion of dosimetry parameters used to calculate tumor doses, to the delivery of the radiation dose. The QA program provided to international institutions includes: 1) remote TLD/OSLD audit of machine output, 2) credentialing for advanced technologies, and 3) review of patient treatment records. IROC Houston uses the same standards and acceptance criteria for all of its audits whether for North American or international sites. Results: IROC Houston’s QA program has reached out to radiotherapy sites in 43 different countries since 2013 through their participation in clinical trials. In the past two years, 2,778 international megavoltage beam outputs were audited with OSLD/TLD. While the average IROC/Inst ratio is near unity for all sites monitored, there are international regions whose results are significantly different from the NA region. In the past 2 years, 477 and 87 IMRT H&N phantoms were irradiated at NA and international sites, respectively. Regardless of the OSLD beam audit results, the overall pass rate (87 percent) for all international sites (no region separation) is equal to the NA sites. Of the 182 international patient charts reviewed, 10.7 percent of the dose calculation points did not meet our acceptance criterion as compared to 13.6 percent for NA sites. The lower pass rate for NA sites results from a much larger brachytherapy component which has been shown to be more error prone. Conclusion: IROC Houston has expanded its QA services worldwide and continues a long history of improving radiotherapy dose delivery in many countries. Funding received for QA audit services from the Korean GOG, DAHANCA, EORTC, ICON and CMIC Group

SU‐E‐T‐187: The Value of Rapid Reviews

PURPOSE The Radiological Physics Center (RPC) performs rapid reviews for several different study groups and for a variety of disease sites including colon, breast, endometrial and cervix. Rapid Reviews have been performed for high dose rate brachytherapy studies, 3D CRT and IMRT studies. The purpose of rapid reviews is to verify that the radiation oncologist is capable of treating a patient per protocol prior to treatment commencing with the goal of reducing the number of deviations. METHODS The rapid review process requires that the institution electronically submit the protocol patient treatment plan prior to the commencement of treatment for a dosimetric and clinical review. Dependent on the protocol, the first patient or every patient submitted by a physician might require a rapid review. Rapid reviews enable the RPC to provide feedback to the physician to rectify errors prior to the start of treatment. Deviations are assessed according to defined criteria within the specific protocolResults: For three protocols, where rapid reviews were required for the first patient placed on protocol, 24%, 48% and 53% required a revision and resubmission for a re-review due to a significant protocol deviation. For one protocol, where rapid reviews were required for all patients, 81% of the submitted patient cases required a revision and resubmission for a re-review. Radiation Oncologists who completed the rapid review process received no major deviations on subsequent patients placed on protocol. CONCLUSION Rapid reviews serve the purpose of reducing the number of protocol deviations by providing feedback to Radiation Oncologists on how to better comply with the requirements of the protocol prior to commencing treatment of a patient on the study.This work was supported by PHS grants CA10953 and CA081647 awarded by NCI, DHHS. This work was supported by PHS grants CA10953 and CA081647 awarded by NCI, DHHS.

SU‐E‐T‐199: The Radiological Physics Center's Credentialing Dosimetry Reviews: Their Effect on Clinical Trial Deviation Rates

PURPOSE To describe the Radiological Physics Centers (RPC) methods to evaluate an institutions ability to meet protocol guidelines in order to decrease NCI clinical trial deviation rate. METHODS The RPCs dosimetry group utilizes 3 methods of assessing an institutions ability to meet the protocol treatment specifications. These methods involve a clinical and dosimetric review of a treatment plan submitted by the institution prior to the first patient being treated on a protocol. The three evaluation methods include use of site/treatment modality specific benchmark cases, evaluation of a previous patient treated in a similar fashion and a rapid review of the first patient placed on a trial prior to start of treatment. The dosimetric review consists of an independent dose recalculation using RPC measured data or RPC standard dosimetry data. The clinical review assesses the patients DVHs and contouring of the tumor volume and critical structures, typically in conjunction with a radiation oncologist. RESULTS Over the past 5 years the RPC has performed these QA reviews for several of the clinical trial groups for several different disease sites and treatment modalities. We have reviewed 1366 treatment plans as a part of credentialing (97 gynecological, 223 prostate, 1046 breast) where 222 failed the first submission requiring the RPC to interact with the submitting institution to resolve the discrepancy. The review of the benchmarks has resulted in 18% of the institutions requiring intervention by the RPC. Performing these reviews has identified potential clinical and dosimetric problem areas that could possibly have resulted in 17% of the charts reviewed to receive a minor or major deviation. CONCLUSIONS The RPCs clinical and dosimetry review of submitted treatment plans before or early in the treatment process has helped to reduce the deviation rates on protocols. Work supported by PHS grant CA 10953 awarded by NCI, DHHS.

SU‐E‐T‐367: Cervix Brachytherapy Dosimetry: Inconsistencies in Defining Bladder and Rectal Points

Purpose: The Radiological Physics Center (RPC) reviews patient records on brachytherapy cervix trials for completeness, consistency with the protocol and dosimetric accuracy to minimize patient dose delivery and reporting uncertainty for NCI funded clinical trials. Within cervical protocols, bladder and rectum doses are specified to limit toxicity to these normal tissues. However, bladder and rectal doses reported by institutions often disagree with the RPCs calculated doses. The RPC has investigated the sources of these disagreements. Methods: The RPC reviewed 182 HDR brachytherapy (tandem and ovoids (T&O)/tandem and ring (T&R)) implants and compared the institutions bladder and rectum point locations and doses to those determined by the RPC strictly adhering to protocol specifications (ICRU 38) for point location and using its independent dose calculation algorithm. The RPC also analyzed its own uncertainty in defining these two points. A dose agreement criterion of ±15% was used as agreed upon by the study group and RPC. Results: The RPC disagreed with the bladder and rectal doses in 25% and 45%, respectively, of the 182 implants. The RPCs own uncertainty in defining the bladder and rectal points was 1mm±0. 1(STDEV), respectively which in a worst case scenario might account for 7% of the dose disagreement. The majority of the dose disagreements were due to the institutions incorrect localization of the bladder and rectal points, by greater than 5mm and 4mm, respectively, away from the ICRU 38 defined location. There were no differences noted whether the applicator used was a T&O or T&R. Conclusions: Most errors resulted from institutions incorrectly defining the bladder and rectum dose calculation points per ICRU 38. Additional education, timely reviews of implant data and communication with institutions are needed to reduce the number of discrepancies. Work supported by PHS grant CA 10953 awarded by NCI, DHHS

SU-E-T-606: Requirements for Performing a Retrospective Patient Chart Review at the Radiological Physics Center (RPC) for Clinical Trials

Purpose: One of the RPCs quality audits used to assure the NCI and Cooperative Trial Groups that institutions participating in clinical trials deliver and report radiation doses that are clinically comparable and consistent is a retrospective review of clinical patient treatment charts. However, there is no standard regarding what patient and dosimetry data to include within a submitted trial patients chart depending on treatment modality (brachytherapy vs. external beam) and protocol specific requirements. This work identifies the required data needed to perform a clinical trial quality audit review based on the evaluation of nearly 2000 patient charts. Methods: Since 2005, the RPC reviewed 1997 patient charts equating to over 13,000 points of calculation. In order to perform these dose recalculations, a minimal amount of data is needed for external beam and brachytherapytreatments. A review of these charts has identified the required patient specific and machine specific data required. In addition the data needs to be submitted in a useable format (CTimages submitted in DICOM format, isodose lines and DVHs in color). Results: Comprehensive data requirements for external beam and brachytherapy are presented. Since 2005, the RPC sent out 1021 letters requesting data or clarifications regarding the treatment. 86% of these requests were for patient specific information. The most common information omitted from a brachytherapy chart were the HDR dwell times and locations, and for external beam charts it was the daily treatment records indicating the monitor units delivered per field. Conclusions: For the RPC to state that trial patient doses are clinically comparable and consistent, the necessary patient and dosimetry data must be submitted in a timely manner. Development of a required data submission checklist to be included with each protocol will minimize trial data submission deficiencies and increase the efficiency of the RPCs quality audits. Work supported by PHS grant CA 10953 awarded by NCI, DHHS

SU‐FF‐T‐25: Cervix Brachytherapy Dosimetry: Observed Improvement in Data Submitted to Clinical Trials

Purpose: For over 40 years, the Radiological Physics Center (RPC) has reviewed the completeness, consistency with protocols, and the dosimetric accuracy of data submitted for cervix patients treated with brachytherapy on national clinical trials. In 2005 the RPC presented data which showed that 65% of cervix patients placed on study had one or more dosimetry errors in at least one of their implants. This study was conducted to determine if brachytherapydosimetry has improved in the last 4 years. Methods & Materials: In the last 4 years the RPC has reviewed 695 HDR and LDR implants. Independent dose calculations were performed at points A, B, vaginal surface, bladder and rectum as defined by the protocol in accordance with ICRU‐38. The vaginal surface dose was defined as a point lateral to the center of the source (s) at 0.5cm from the surface of the ovoid. RPC doses were compared to the institutions reported doses.Results: From 2005 to 2009 the RPC has determined that dosimetry errors have decreased by 50%. Most remaining errors result from incorrectly defining calculation points. However the use of CT has created some new issues due to the manipulation of the CTimages when visualizing the tandem. Conclusion: The decrease in dosimetry errors is due to several different aspects: increase in rapid reviews, more timely reviews of patients and better communication with an institution when discrepancies are discovered. This work was supported by PHS grant CA 10953 awarded by NCI, DHHS

SU‐FF‐T‐410: The Credentialing Process for the NSABP B‐39/RTOG 0413 Partial Breast Irradiation Trial

Purpose: Develop a credentialing process for the NSABP B‐39/RTOG 0413 Partial Breast Irradiation (PBI) trial. Method: NSABP B‐39/RTOG 0413, a Phase III trial comparing whole breast irradiation versus PBI (3D conformal radiation therapy (3D‐CRT), MammoSite® and multi‐catheter brachytherapy). For each PBI technique, an institution, radiationoncologist and physicist team must be credentialed. The credentialing verifies that all personnel involved with treatment planning have read the protocol prior to enrolling patients to limit the number of deviations. Credentialing also allows feedback to the team prior to patient treatment to correct any mistakes. Each institution must complete online the knowledge assessment and facility questionnaires and download a PBI technique CT benchmark case. Results: Teams at 308 distinct institutions have submitted applications for credentialing for at least one PBI technique. 733 radiationoncologists applied for 3D‐CRT credentialing, 490 radiationoncologists for MammoSite® and 151 radiationoncologists for multicatheter. Of those applications, 79% became credentialed for 3D‐CRT, 69% for MammoSite®, and 56% for multi‐catheter. Reasons for which a radiationoncologist failed to become credentialed included; incomplete application, incorrect answers on knowledge assessment, treatment planning system could not submit data electronically, and the CT benchmark was not planned per protocol. The first patient enrolled by each institution received a rapid review prior to patient treatment. The next 4 cases received a timely review. These reviews included a dosimetric and clinical review. Currently, this protocol has accrued 880 patients, of which 328 treated with 3D‐CRT, 82 treated with MammoSite, and 31 treated with multi‐catheter brachytherapy. Of the 441 patients treated to date on the PBI arm there have been no dosimetric deviations. Conclusion: The PBI credentialing process has been successful in educating participating facilities and helping to minimize dosimetry errors. Work was supported by PHS grants CA10953 and CA081647 awarded by NCI.