- Email: [email protected]
Quality control of radiation therapy in multi-institutional randomized clinical trial for localized prostate cancer
QUALITY CONTROL
OF RADIATION THERAPY IN
MULTI-INSTITUTIONAL
RANDOMIZED
TRIAL FOR LOCALIZED
PROSTATE CANCER
MARK D. HAFERMANN,
CLINICAL
M.D.
ROBERT P GIBBONS, M.D. GERALD
I? MURPHY, M.D., D.Sc.
INVESTIGATORS OF NATIONALPROSTATICCANCERPROJECT* From the Sections of Radiation Oncology and Urology, Virginia Mason Medical Center, Seattle, Washington, and the Department of Urology, State University of New York at Buffalo, New York
ABSTRACT-The National Prostatic Cancer Project (NPCP) f ram 1978 through 1985 compared definitive radiation therapy for Stages Bz, C, D1 lesions in those who received only radiation treatment to those who received two years of additional cyclophosphamide (Cytoxan) or estramustine phosphate (Emcyt) chemotherapy. Two hundred fifty-four patients were entered and 229 evaluated for compliance of the spatial localization of the prostate through review of the simulation and port films. In 78 per cent this was satisfactory, whereas in 12 per cent it was unsatisfactory, and another 10 per cent were not evaluable. The principle cause of an unsatisfactory rating was failure to adequately cover the prostatic target volume, especially the apex which was found to be variable in location. Routine use of retrograde urethrocystography is urged as part of the localization method in patients to receive definitive external beam radiation therapy for prostate cancer. The role and impact of quality assurance programs for radiotherapy in cooperative clinical study groups is reviewed and discussed.
In May, 1978, the National Prostatic Cancer Project (NPCP) activated Protocol 1000 which compared, in pelvic lymph node-staged disease (surgical Stages Be, C, Di), radiotherapy (6,600-7,020 rad/33-39 fractions15 weeks) with no added treatment versus two years of either cyclophosphamide (Cytoxan) or estramustine
phosphate sodium (Emcyt) .l Accrual of patients was terminated October, 1985, and data were evaluated as recently as July, 1986. Initial quality assurance efforts in this study led to the recognition of considerable variation in localization techniques and parameters among the participating institutions despite
*Other contributors: Robert Huhen, M.D., Roswell Park Memorial Institute, Buffalo, New York; Raju Thomas, M.D., ‘Iblane University, New Orleans, Louisiana; Stefan A. Loening, M.D., Universitv of Iowa Hospitals & Clinics, Iowa City, Iowa; David G. McLeod, M.D., Walter Reed Army Medical Center, Washington, D.C.; Jean B. deKernion, M.D., University of California, Los Angeles, California; Chirpriya Dhabuwala, M.D., Wayne State University School of Medicine, Detroit, Michigan; Peter Scardino, M.D., Baylor College of Medicine, Houston,
Texas; William W. Scott, M.D., Johns Hopkins Hospital, Baltimore, Maryland; Mark Soloway, M.D., University of Tennessee, Memphis, Tennessee; Patrick Guinan, M.D., Cook County Hospital, Chicago, Illinois; John Lynch, M.D., Georgetown University, Washington, D.C.; Robert Flardgan, M.D., University of Kentucky, Lexington, Kentucky; Brian Miles, M.D., Henry Ford Hospital, Detroit, Michigan; Hugh Fisher, M.D., Albany Medical College Hospital, Albany, New York; John lkachtenberg, M.D., University of Toronto General Hospital, Toronto, Canada.
UROLOGY
I
FEBRUARY
1988
/
VOLUME
XxX1,
NUMBER
2
119
FIGURE 1. (A, B) Orthogonal pair of simulation radiographs illustrating urethrocystogram _ prostate boost fields out&ed.
recommended guidelines in the protocol. Special efforts were subsequently made to assure that the radiotherapy was delivered correctly, with particular consideration given to localization of the prostatic volume and prostatic apex. Quality control of radiotherapy in clinical trials is a current and necessary requirement which is appreciated on a worldwide basis.2 We believe that an experience in this regard should be shared by this report. Material and Methods All patients receiving external beam irradiation in this NPCP protocol underwent a localization and treatment planning procedure. At the outset of the study, the protocol called for a commonly used method of localization of the prostatic volume. 3 Through an indwelling urethral catheter, a water-soluble iodinated contrast agent was instilled in the bladder and, through another catheter, barium was injected into the rectum, opacifing these organs on two pairs of orthogonal radiographs (simulation films). When indicated, one pair was used to outline the pelvic fields for treatment of the prostate plus the regional lymph nodes. The other pair of radiographs was used to determine the target volume for the prostatic boost fields. The location of the prostate was deduced from anatomic landmarks consisting of: the symphysis pubis, anteriorly; the base of the bladder, superiorly; and the anterior rectal wall, posteriorly. The apex of the prostate is not specifically identified by this method, but the inferior limit of the field was to be a line drawn
120
with pelvic and
between the ischial tuberosities, which was considered to provide an adequate margin. Copies of the simulation and portal check films for both the pelvic and prostate-only fields were requested to be sent to the NPCP radiotherapy review office at the Virginia Mason Medical Center, Seattle, Washington, where they were reviewed for compliance with protocol standards. Periodic reports were presented to the participating institutions from 1978 to May, 1986. This report completes the analysis of compliance in terms of the spatial localization of the pelvic fields and the prostatic volume, as the protocol required. Dosimetric review is done by the Radiological Physics Center, Houston, Texas, and is not covered in this report. Results Our experience in this study for compliance to localization criteria of the pelvic and prostate volumes for definitive external beam irradiation among the 15 participating institutions showed, of 254 patients accessioned into Protocol 1000, 18 patients received Iodine-125 interstitial therapy and 7 patients withdrew without treatment, leaving 229 patients who received external beam irradiation. In 13 of the 229 cases (6%) films were not received for review, leaving 216 cases (94%) for evaluation of the adequacy of radiation field localization. Of the 216 cases, 169 or 78 per cent were judged satisfactory, 25 or 12 per cent were judged unsatisfactory, and 22 or 10 per cent were inevaluable for a number of reasons including technically inadequate films which made it impossible to
UROLOGY
/ FEBRUARY
1988
/ VOLUME
XXX.
NUMBER 2
FIGURE 2. (A, B, C) These 3 cases illustrate range of location of membranous urethra and urogenital diaphragm as they project under symphysis pubis. Setting inferior margin of radiation field on a line connecting ischial tuberosities provides adequate coverage in majority of patients-but in some cases this margin is unnecessarily excessive and in a few cases this margin is inadequate. Technique does not rely on arbitrary anatomic landmarks and eliminates uncertainty with respect to proper placement of inferior margin of field.
make a judgment. If the inevaluable cases are excluded from the denominator, the satisfactory rate is 87 per cent and the unsatisfactory rate is 13 per cent. The early film reviews revealed approximately a 25 per cent rate of noncompliance with the field definitions of the protocol. A common problem was concern over adequate coverage of the inferior margin of the prostate on the small coned-down boost fields. Although the criteria set forth in the protocol provided adequate coverage of the prostatic apex in the majority of cases, it became apparent, on review, that there was a certain amount of inconsistency in the placement of the inferior margin of the boost fields by various radiation oncologists. Accordingly, a revised technique for localization which employs a retrograde urethral injection of contrast material (thereby obtaining a urethrocystogram) was devised (Fig. l).” The procedure identifies the urogenital diaphragm and thus delineates the location of the apex of the prostate on the localization radiographs. This information is not obtained if the bladder is filled with contrast agent through a ureteral catheter. It has been found that the level of the urogenital diaphragm, as it projects below the subpubic arch, varies to a modest degree from patient to patient (Fig. 2). The modified localization procedure eliminates the uncertainty with regard to proper placement of the inferior margin and provides a more custom-made field for the individual patient. Moreover, the injection technique is simpler, less traumatic, and requires less expensive ma-
UROLOGY
/ FEBRUARY 1988 I VOLUMEXXXLNUMBERZ
terials than the original catheter technique. Computerized tomography (CT) does not define the apex of the gland as well as the urethrogram does since there is no delineation between the prostatic tissue and the urogenital diaphragm. On the other hand, CT is helpful in defining the superior extent and/or lateral shape and dimensions of the prostatic neoplasm when there is palpable seminal vesicle involvement or a very large prostate. Save for the exceptions noted, CT provides no localization advantage over retrograde urethrocystography and standard simulation films in the majority of patients, but it does add additional expense. The revised technique was introduced and discussed through two workshops and written communication during the course of the study. The final analyses showed a decrease of approximately 50 per cent in the rate of unacceptable localizations. Comment Therapy of prostate cancer has specific proponents favoring surgery or radiotherapy for localized diseases5 This report does not address this issue but many believe that multicenter randomized trials are still required to decide this question, as well as others involving the use of radiotherapy. It is recognized that acceptable variations in treatment policies exist among various institutions and radiation oncologists. In addition, therapy units, simulation equipment, and procedures for localization and documentation of the target volume differ. Nevertheless, in the
121
context of a clinical study it is essential to establish uniformity and consistency of treatment if the analysis of end results is to be meaningful. One assumes that at centers of excellence, the treatment delivered is well conceived, given in the prescribed manner, and properly monitored. As other cooperative groups have found, this is not always the case and requires continuous quality control monitoring.e Modern radiation therapy equipment allows the delivery of radiation to a predetermined tumor volume with considerable precision, providing care is taken in the initial localization of the target volume and day-to-day set up of the radiation fields. Radiation therapy simulators provide a facile means of localizing the tumor volume by radiographic means and indexing that volume to a mark on the patient’s skin. Subsequently, radiographs of the treatment fields (port films) exposed by the actual treatment beam, are made to verify that the simulated treatment fields are appropriately reproduced by the therapy unit. Computer-based treatment planning systems provide accurate information on the dose distribution throughout the tumor volume and normal tissues. (As indicated by Cunningham,’ the target for accuracy in radiotherapy can reasonably be set at f 5 % . The precision of the three separate procedures in dose delivery are of:2.5 % for dose calibration, + 3-4 % for relative dose calculations, and an estimated + 3-4% for dose delivery methods. Assuming these uncertainties occur at random, in either direction, the overall uncertainty is 4.9-6.2% .“) These techniques verify and document, for current and retrospective review, the target volume treated and the dose of radiation delivered, thereby providing an excellent method of quality control. In general, quality reviews of radiation therapy are divided into two components-an assessment of the accuracy and adequateness of (1) treatment dosimetry and (2) spatial localization of the treatment volume through an assessment of the simulation and port films. Although some objective criteria can be defined in scoring localization, the assessment is, to a certain degree, dependent on the subjective interpretation of the reviewer; whereas, dosimetry is assessed by objective numerical data. This report deals with the simulator and port film review by a single person (MDH) and does not consider the dosimetric data, which are reviewed by the Radiological Physics Center (RPC), and remain to be analyzed. Finding
122
that 12 per cent of the cases reviewed failed to comply with protocol standards and another 10 per cent could not be evaluated adequately due to lack of proper films or poor quality films leaves 78 per cent of cases considered in compliance and raises a number of questions regarding the practicalities of carrying out multiinstitutional studies, the validity of data from these studies, and the overall quality of radiotherapy practiced. An attempt to answer these questions, or at least place the questions and these results in perspective, will follow. It should be noted that during the early and middle years of patient accrual to Protocol 1000 there was only a 70-80 per cent compliance in the submission of films and data, the inevaluable rate was 38 per cent, and unsatisfactory rate approximately 25 per cent. After concerted effort and two quality assurance workshops the submission rate rose to 94 per cent, the inevaluable rate dropped to 10 per cent and the unsatisfactory rate to 12 per cent. This experience is not dissimilar from other cooperative groups carrying out multi-institutional clinical trials. 68,g The final data on compliance can be refined further. It is not unreasonable to assume that the ratio of satisfactory versus unsatisfactory localizations in the inevaluable group is roughly the same as in the patient group as a whole. This assumption allows us to eliminate the inevaluable cases from the denominator. Thus, considering only evaluable cases, the satisfactory rate becomes 87 per cent and unsatisfactory rate is 13 per cent. These figures fall within the guidelines for credibility suggested by Davis,lO who states the practical goal of quality assurance measures is to reduce the noneligibility and noncompliance rates to =5 per cent, but should not exceed 20 per cent if a study’s credibility is to be maintained. To put the NPCP’s experience with quality control of radiation therapy into further perspective it’is helpful to examine the experience of other multi-institutional cooperative groups. The pioneering efforts with respect to quality control of radiation therapy in clinical trials and subsequent research initiatives in quality control were made by the Radiotherapy Committee of the Cancer and Leukemia Group B (CALGB) .6 The Quality Assurance Review Center (QUARC) is an outgrowth of the CALGB quality assurance program and now serves multiple national and international cooperative groups with interventional review
UROLOGY
i
FEBRUARY 1988
I
VOLUME XXXI. NUMBER 2
processes.” In the CALGB, the impact of the quality assurance program on both data accrual and compliance is noted by the fact that at the time quality assurance measures were implemented 75 per cent of the data was not available for analysis. Within three months, about 50 per cent of the data was available with a deviation rate of approximately 40 per cent. By two years, 90 per cent of the data was available and the deviation rate had decreased to <5 per cent. 6 QUARC examined the major protocol deviations in 2,258 cases treated on eighteen different protocols. They found, among other parameters, that the initial major deviation rate for field review was 23 per cent. With the implementation of a feedback mechanism to the radiotherapist this major deviation rate dropped progressively to under 5 per cent after a participation time of thirty-one months, or more.” There are two types of quality assurance reviews: (1) The “final review assessment” in which the case materials are reviewed retrospectively after the treatment of a given patient is completed. While this process usually goes on during the course of the study, it is not timely enough to affect the treatment of a given patient. (2) The “interventional mode” or “on-treatment review” in which the case materials are reviewed within first week of radiation treatment and provides an interactive mechanism to correct or compensate for deviations before they unalterably affect the final dose distribution.” QUARC found that on treatment reviews reduced the major deviation rate on treatment volume in a group of 696 patients treated on multiple protocols from 20 to 10 per cent.” It is believed there is educational value from participation in cooperative studies employing quality assurance methods which tend to improve the overall quality of treatment within an institution. However, the major objective of the program from the point of view of the clinical trial is to increase the evaluability rate of patients accessioned to the studies. The additional efforts are believed to be cost-effective.e Although it is too soon to report therapeutic results in protocol 1000, it should be possible to analyze outcome data separately for the categories of compliance or noncompliance. Providing this group of patients can be properly followed over a longer time, an opportunity exists to learn more regarding dose response relationships and perhaps the impact of local control of prostate cancer on survival.
UROLOGY
/
FEBRUARY 1988
I
The significance of marginal underdoses or geographic misses has been underscored by several reports involving different disease sites. In the Patterns of Care Study, Hanks and Kramer12 found a significant correlation between the dose in the prostate (p = 0.04), and in the paraprostatic tissues (p = O.Ol), with increased local recurrence rates, indicating that marginal underdoses, as well as central prostatic underdosage, produced local recurrences. In the Hodgkin disease review it was found that if a port film indicated a portion of the disease was not adequately covered by the field, the infieldmarginal recurrence rate was 33 per cent; whereas, if the disease was adequately encompassed, the infield-marginal recurrence rate was 7 per cent.12 Perez, Gardner, and Glasgow,9 citing a clinical trial for patients with small cell carcinoma of the lung, described an increased incidence of local failures associated with inadequate radiation fields compared with adequate fields. In another clinical trial for patients with non-small cell cancer of the lung, better local tumor control and survival was seen for the patient group irradiated in full compliance or only minor protocol variations compared with the patient group with major deviations.e What are the implications of these quality control analyses with respect to the general community practice of radiation oncology? While it is acknowledged that the quantifiable aspects for radiation dose delivery lends itself to quality control, perhaps more than other cancer treatment modalities, radiation oncology, as a clinical discipline, is leading the way in self analyses as exemplified by the national Patterns of Care Study (PCS). This effort was established by the American College of Radiology in 1973 and is supported by the National Cancer Institute to assess and document the status of radiation therapy with respect to its quality and accessibility. 13-15The project examines the structure, processes, and outcome of care of radiation therapy as practiced across a broad front in the United States. On the whole, the quality of radiotherapeutic practice, based on outcome analyses, was found to be very good.‘5 Beginning in 1985, under the auspices of the standing Quality Assessment Committees of the American College of Radiology and the American Society for Therapeutic Radiology and Oncology, the PCS began offering a voluntary quality assessment service to individual facilities with the objectives of further improving the
VOLUME XxX1, NUMBER 2
123
quality of care at the institutional level and conduct quality assessment research. l6 The Patterns of Care Study for Radiation Oncology is the first nationwide evaluation of a clinical discipline and provides a model for other oncology specialties. Department of Urology State University of New York at Buffalo 139 Parker Hall Buffalo, New York 14214 (DR. MURPHY) References 1. Murphy GP: Current status of the National Prostatic Cancer Project treatment protocols, in Denis L, Murphy GP, Prout GR, and Schroder F (Eds): Controlled Clinical ‘IHals in Urologic Oncology, New York, Raven Press, 1984, vol. 13, pp 119-133. 2. Quality Assurance in Radiation Therapy: Clinical and physical aspects. Proceedings of the First Intrnational Symposium on Quality Assessment in Radiation Oncology, Int J Radiat Oncol Biol Phys 10 (suppl. 1): (1984). 3. Gardner A, Bagshaw MA, Page V, and Karzmark CJ: Tumor localization, dosimetry, simulation and treatment procedures in radiotherapy: the isocentric technique, AJR 114: 163 (1972). 4. Hafermann MD: Cancer of the prostate-external radiotherapy, in Murphy GP (Ed): Clinics in Oncology, London,
124
WB Saunders Company Ltd, 1983, vol. 212, pp 371-405. 5. Huben Rp, and Murphy GP: Prostate cancer: an update, CA 36: 274 (1986). 6. Clicksman AS, Reinstein LE, and Laurie F: Quality assurance of radiotherapy in clinical trials, Cancer Treat Rep 69: 1199 (1985). 7. Cunningham JR: Quality assurance in dosimetry and treatment planning, Int J Radiat Oncol Biol Phys lO(supp1. 1): 105 (1984). 8. Glicksman AS, Reinstein LE, MC Shan D, and Laurie F: Radiotherapy quality assurance program in a cooperative group, ibid 7: 1561 (1981). 9. Perez CA, Gardner P, and Glasgow GP: Radiotherapy quality assurance in clinical trials, ibid lO(supp1. 1): 119 (1984). ibid 10. Davis LW: Quality assurance in clinical trials, lO(supp1. 1): 147 (1984). 11. Reinstein LE, Peachey BS, Laurie BS, and Glicksman AS: Impact of a dosimetry review program on radiotherapy in group trials, ibid 11:1179 (1985). 12. Hanks GE, and Kramer S: Consensus of best current management: the starting point for clinical quality assessment, ibid lO(supp1. 1): 87 (1984). 13. Kramer S: The study of the Patterns of Care in Radiation Therapy, Cancer 39: 780 (1977). 14. Kramer S, and Herring D: The Patterns of Care Study: a nationwide evaluation of the practice of radiation therapy in cancer management, Int J Radiat Oncol Biol Phys 1: 1231 (1976). 15. Kramer S: The patterns of clinical care in radiation therapy in the United States, ibid lO(supp1. 1): 49 (1984). 16. Hanks GE: Future plans for quality assurance in radiation oncology in the United States, ibid lO(supp1. 1): 35 (1984).
UROLOGY
/ FEBRUARY
1988
/ VOLUME
XXKI, NUMBER
2
OF RADIATION THERAPY IN
MULTI-INSTITUTIONAL
RANDOMIZED
TRIAL FOR LOCALIZED
PROSTATE CANCER
MARK D. HAFERMANN,
CLINICAL
M.D.
ROBERT P GIBBONS, M.D. GERALD
I? MURPHY, M.D., D.Sc.
INVESTIGATORS OF NATIONALPROSTATICCANCERPROJECT* From the Sections of Radiation Oncology and Urology, Virginia Mason Medical Center, Seattle, Washington, and the Department of Urology, State University of New York at Buffalo, New York
ABSTRACT-The National Prostatic Cancer Project (NPCP) f ram 1978 through 1985 compared definitive radiation therapy for Stages Bz, C, D1 lesions in those who received only radiation treatment to those who received two years of additional cyclophosphamide (Cytoxan) or estramustine phosphate (Emcyt) chemotherapy. Two hundred fifty-four patients were entered and 229 evaluated for compliance of the spatial localization of the prostate through review of the simulation and port films. In 78 per cent this was satisfactory, whereas in 12 per cent it was unsatisfactory, and another 10 per cent were not evaluable. The principle cause of an unsatisfactory rating was failure to adequately cover the prostatic target volume, especially the apex which was found to be variable in location. Routine use of retrograde urethrocystography is urged as part of the localization method in patients to receive definitive external beam radiation therapy for prostate cancer. The role and impact of quality assurance programs for radiotherapy in cooperative clinical study groups is reviewed and discussed.
In May, 1978, the National Prostatic Cancer Project (NPCP) activated Protocol 1000 which compared, in pelvic lymph node-staged disease (surgical Stages Be, C, Di), radiotherapy (6,600-7,020 rad/33-39 fractions15 weeks) with no added treatment versus two years of either cyclophosphamide (Cytoxan) or estramustine
phosphate sodium (Emcyt) .l Accrual of patients was terminated October, 1985, and data were evaluated as recently as July, 1986. Initial quality assurance efforts in this study led to the recognition of considerable variation in localization techniques and parameters among the participating institutions despite
*Other contributors: Robert Huhen, M.D., Roswell Park Memorial Institute, Buffalo, New York; Raju Thomas, M.D., ‘Iblane University, New Orleans, Louisiana; Stefan A. Loening, M.D., Universitv of Iowa Hospitals & Clinics, Iowa City, Iowa; David G. McLeod, M.D., Walter Reed Army Medical Center, Washington, D.C.; Jean B. deKernion, M.D., University of California, Los Angeles, California; Chirpriya Dhabuwala, M.D., Wayne State University School of Medicine, Detroit, Michigan; Peter Scardino, M.D., Baylor College of Medicine, Houston,
Texas; William W. Scott, M.D., Johns Hopkins Hospital, Baltimore, Maryland; Mark Soloway, M.D., University of Tennessee, Memphis, Tennessee; Patrick Guinan, M.D., Cook County Hospital, Chicago, Illinois; John Lynch, M.D., Georgetown University, Washington, D.C.; Robert Flardgan, M.D., University of Kentucky, Lexington, Kentucky; Brian Miles, M.D., Henry Ford Hospital, Detroit, Michigan; Hugh Fisher, M.D., Albany Medical College Hospital, Albany, New York; John lkachtenberg, M.D., University of Toronto General Hospital, Toronto, Canada.
UROLOGY
I
FEBRUARY
1988
/
VOLUME
XxX1,
NUMBER
2
119
FIGURE 1. (A, B) Orthogonal pair of simulation radiographs illustrating urethrocystogram _ prostate boost fields out&ed.
recommended guidelines in the protocol. Special efforts were subsequently made to assure that the radiotherapy was delivered correctly, with particular consideration given to localization of the prostatic volume and prostatic apex. Quality control of radiotherapy in clinical trials is a current and necessary requirement which is appreciated on a worldwide basis.2 We believe that an experience in this regard should be shared by this report. Material and Methods All patients receiving external beam irradiation in this NPCP protocol underwent a localization and treatment planning procedure. At the outset of the study, the protocol called for a commonly used method of localization of the prostatic volume. 3 Through an indwelling urethral catheter, a water-soluble iodinated contrast agent was instilled in the bladder and, through another catheter, barium was injected into the rectum, opacifing these organs on two pairs of orthogonal radiographs (simulation films). When indicated, one pair was used to outline the pelvic fields for treatment of the prostate plus the regional lymph nodes. The other pair of radiographs was used to determine the target volume for the prostatic boost fields. The location of the prostate was deduced from anatomic landmarks consisting of: the symphysis pubis, anteriorly; the base of the bladder, superiorly; and the anterior rectal wall, posteriorly. The apex of the prostate is not specifically identified by this method, but the inferior limit of the field was to be a line drawn
120
with pelvic and
between the ischial tuberosities, which was considered to provide an adequate margin. Copies of the simulation and portal check films for both the pelvic and prostate-only fields were requested to be sent to the NPCP radiotherapy review office at the Virginia Mason Medical Center, Seattle, Washington, where they were reviewed for compliance with protocol standards. Periodic reports were presented to the participating institutions from 1978 to May, 1986. This report completes the analysis of compliance in terms of the spatial localization of the pelvic fields and the prostatic volume, as the protocol required. Dosimetric review is done by the Radiological Physics Center, Houston, Texas, and is not covered in this report. Results Our experience in this study for compliance to localization criteria of the pelvic and prostate volumes for definitive external beam irradiation among the 15 participating institutions showed, of 254 patients accessioned into Protocol 1000, 18 patients received Iodine-125 interstitial therapy and 7 patients withdrew without treatment, leaving 229 patients who received external beam irradiation. In 13 of the 229 cases (6%) films were not received for review, leaving 216 cases (94%) for evaluation of the adequacy of radiation field localization. Of the 216 cases, 169 or 78 per cent were judged satisfactory, 25 or 12 per cent were judged unsatisfactory, and 22 or 10 per cent were inevaluable for a number of reasons including technically inadequate films which made it impossible to
UROLOGY
/ FEBRUARY
1988
/ VOLUME
XXX.
NUMBER 2
FIGURE 2. (A, B, C) These 3 cases illustrate range of location of membranous urethra and urogenital diaphragm as they project under symphysis pubis. Setting inferior margin of radiation field on a line connecting ischial tuberosities provides adequate coverage in majority of patients-but in some cases this margin is unnecessarily excessive and in a few cases this margin is inadequate. Technique does not rely on arbitrary anatomic landmarks and eliminates uncertainty with respect to proper placement of inferior margin of field.
make a judgment. If the inevaluable cases are excluded from the denominator, the satisfactory rate is 87 per cent and the unsatisfactory rate is 13 per cent. The early film reviews revealed approximately a 25 per cent rate of noncompliance with the field definitions of the protocol. A common problem was concern over adequate coverage of the inferior margin of the prostate on the small coned-down boost fields. Although the criteria set forth in the protocol provided adequate coverage of the prostatic apex in the majority of cases, it became apparent, on review, that there was a certain amount of inconsistency in the placement of the inferior margin of the boost fields by various radiation oncologists. Accordingly, a revised technique for localization which employs a retrograde urethral injection of contrast material (thereby obtaining a urethrocystogram) was devised (Fig. l).” The procedure identifies the urogenital diaphragm and thus delineates the location of the apex of the prostate on the localization radiographs. This information is not obtained if the bladder is filled with contrast agent through a ureteral catheter. It has been found that the level of the urogenital diaphragm, as it projects below the subpubic arch, varies to a modest degree from patient to patient (Fig. 2). The modified localization procedure eliminates the uncertainty with regard to proper placement of the inferior margin and provides a more custom-made field for the individual patient. Moreover, the injection technique is simpler, less traumatic, and requires less expensive ma-
UROLOGY
/ FEBRUARY 1988 I VOLUMEXXXLNUMBERZ
terials than the original catheter technique. Computerized tomography (CT) does not define the apex of the gland as well as the urethrogram does since there is no delineation between the prostatic tissue and the urogenital diaphragm. On the other hand, CT is helpful in defining the superior extent and/or lateral shape and dimensions of the prostatic neoplasm when there is palpable seminal vesicle involvement or a very large prostate. Save for the exceptions noted, CT provides no localization advantage over retrograde urethrocystography and standard simulation films in the majority of patients, but it does add additional expense. The revised technique was introduced and discussed through two workshops and written communication during the course of the study. The final analyses showed a decrease of approximately 50 per cent in the rate of unacceptable localizations. Comment Therapy of prostate cancer has specific proponents favoring surgery or radiotherapy for localized diseases5 This report does not address this issue but many believe that multicenter randomized trials are still required to decide this question, as well as others involving the use of radiotherapy. It is recognized that acceptable variations in treatment policies exist among various institutions and radiation oncologists. In addition, therapy units, simulation equipment, and procedures for localization and documentation of the target volume differ. Nevertheless, in the
121
context of a clinical study it is essential to establish uniformity and consistency of treatment if the analysis of end results is to be meaningful. One assumes that at centers of excellence, the treatment delivered is well conceived, given in the prescribed manner, and properly monitored. As other cooperative groups have found, this is not always the case and requires continuous quality control monitoring.e Modern radiation therapy equipment allows the delivery of radiation to a predetermined tumor volume with considerable precision, providing care is taken in the initial localization of the target volume and day-to-day set up of the radiation fields. Radiation therapy simulators provide a facile means of localizing the tumor volume by radiographic means and indexing that volume to a mark on the patient’s skin. Subsequently, radiographs of the treatment fields (port films) exposed by the actual treatment beam, are made to verify that the simulated treatment fields are appropriately reproduced by the therapy unit. Computer-based treatment planning systems provide accurate information on the dose distribution throughout the tumor volume and normal tissues. (As indicated by Cunningham,’ the target for accuracy in radiotherapy can reasonably be set at f 5 % . The precision of the three separate procedures in dose delivery are of:2.5 % for dose calibration, + 3-4 % for relative dose calculations, and an estimated + 3-4% for dose delivery methods. Assuming these uncertainties occur at random, in either direction, the overall uncertainty is 4.9-6.2% .“) These techniques verify and document, for current and retrospective review, the target volume treated and the dose of radiation delivered, thereby providing an excellent method of quality control. In general, quality reviews of radiation therapy are divided into two components-an assessment of the accuracy and adequateness of (1) treatment dosimetry and (2) spatial localization of the treatment volume through an assessment of the simulation and port films. Although some objective criteria can be defined in scoring localization, the assessment is, to a certain degree, dependent on the subjective interpretation of the reviewer; whereas, dosimetry is assessed by objective numerical data. This report deals with the simulator and port film review by a single person (MDH) and does not consider the dosimetric data, which are reviewed by the Radiological Physics Center (RPC), and remain to be analyzed. Finding
122
that 12 per cent of the cases reviewed failed to comply with protocol standards and another 10 per cent could not be evaluated adequately due to lack of proper films or poor quality films leaves 78 per cent of cases considered in compliance and raises a number of questions regarding the practicalities of carrying out multiinstitutional studies, the validity of data from these studies, and the overall quality of radiotherapy practiced. An attempt to answer these questions, or at least place the questions and these results in perspective, will follow. It should be noted that during the early and middle years of patient accrual to Protocol 1000 there was only a 70-80 per cent compliance in the submission of films and data, the inevaluable rate was 38 per cent, and unsatisfactory rate approximately 25 per cent. After concerted effort and two quality assurance workshops the submission rate rose to 94 per cent, the inevaluable rate dropped to 10 per cent and the unsatisfactory rate to 12 per cent. This experience is not dissimilar from other cooperative groups carrying out multi-institutional clinical trials. 68,g The final data on compliance can be refined further. It is not unreasonable to assume that the ratio of satisfactory versus unsatisfactory localizations in the inevaluable group is roughly the same as in the patient group as a whole. This assumption allows us to eliminate the inevaluable cases from the denominator. Thus, considering only evaluable cases, the satisfactory rate becomes 87 per cent and unsatisfactory rate is 13 per cent. These figures fall within the guidelines for credibility suggested by Davis,lO who states the practical goal of quality assurance measures is to reduce the noneligibility and noncompliance rates to =5 per cent, but should not exceed 20 per cent if a study’s credibility is to be maintained. To put the NPCP’s experience with quality control of radiation therapy into further perspective it’is helpful to examine the experience of other multi-institutional cooperative groups. The pioneering efforts with respect to quality control of radiation therapy in clinical trials and subsequent research initiatives in quality control were made by the Radiotherapy Committee of the Cancer and Leukemia Group B (CALGB) .6 The Quality Assurance Review Center (QUARC) is an outgrowth of the CALGB quality assurance program and now serves multiple national and international cooperative groups with interventional review
UROLOGY
i
FEBRUARY 1988
I
VOLUME XXXI. NUMBER 2
processes.” In the CALGB, the impact of the quality assurance program on both data accrual and compliance is noted by the fact that at the time quality assurance measures were implemented 75 per cent of the data was not available for analysis. Within three months, about 50 per cent of the data was available with a deviation rate of approximately 40 per cent. By two years, 90 per cent of the data was available and the deviation rate had decreased to <5 per cent. 6 QUARC examined the major protocol deviations in 2,258 cases treated on eighteen different protocols. They found, among other parameters, that the initial major deviation rate for field review was 23 per cent. With the implementation of a feedback mechanism to the radiotherapist this major deviation rate dropped progressively to under 5 per cent after a participation time of thirty-one months, or more.” There are two types of quality assurance reviews: (1) The “final review assessment” in which the case materials are reviewed retrospectively after the treatment of a given patient is completed. While this process usually goes on during the course of the study, it is not timely enough to affect the treatment of a given patient. (2) The “interventional mode” or “on-treatment review” in which the case materials are reviewed within first week of radiation treatment and provides an interactive mechanism to correct or compensate for deviations before they unalterably affect the final dose distribution.” QUARC found that on treatment reviews reduced the major deviation rate on treatment volume in a group of 696 patients treated on multiple protocols from 20 to 10 per cent.” It is believed there is educational value from participation in cooperative studies employing quality assurance methods which tend to improve the overall quality of treatment within an institution. However, the major objective of the program from the point of view of the clinical trial is to increase the evaluability rate of patients accessioned to the studies. The additional efforts are believed to be cost-effective.e Although it is too soon to report therapeutic results in protocol 1000, it should be possible to analyze outcome data separately for the categories of compliance or noncompliance. Providing this group of patients can be properly followed over a longer time, an opportunity exists to learn more regarding dose response relationships and perhaps the impact of local control of prostate cancer on survival.
UROLOGY
/
FEBRUARY 1988
I
The significance of marginal underdoses or geographic misses has been underscored by several reports involving different disease sites. In the Patterns of Care Study, Hanks and Kramer12 found a significant correlation between the dose in the prostate (p = 0.04), and in the paraprostatic tissues (p = O.Ol), with increased local recurrence rates, indicating that marginal underdoses, as well as central prostatic underdosage, produced local recurrences. In the Hodgkin disease review it was found that if a port film indicated a portion of the disease was not adequately covered by the field, the infieldmarginal recurrence rate was 33 per cent; whereas, if the disease was adequately encompassed, the infield-marginal recurrence rate was 7 per cent.12 Perez, Gardner, and Glasgow,9 citing a clinical trial for patients with small cell carcinoma of the lung, described an increased incidence of local failures associated with inadequate radiation fields compared with adequate fields. In another clinical trial for patients with non-small cell cancer of the lung, better local tumor control and survival was seen for the patient group irradiated in full compliance or only minor protocol variations compared with the patient group with major deviations.e What are the implications of these quality control analyses with respect to the general community practice of radiation oncology? While it is acknowledged that the quantifiable aspects for radiation dose delivery lends itself to quality control, perhaps more than other cancer treatment modalities, radiation oncology, as a clinical discipline, is leading the way in self analyses as exemplified by the national Patterns of Care Study (PCS). This effort was established by the American College of Radiology in 1973 and is supported by the National Cancer Institute to assess and document the status of radiation therapy with respect to its quality and accessibility. 13-15The project examines the structure, processes, and outcome of care of radiation therapy as practiced across a broad front in the United States. On the whole, the quality of radiotherapeutic practice, based on outcome analyses, was found to be very good.‘5 Beginning in 1985, under the auspices of the standing Quality Assessment Committees of the American College of Radiology and the American Society for Therapeutic Radiology and Oncology, the PCS began offering a voluntary quality assessment service to individual facilities with the objectives of further improving the
VOLUME XxX1, NUMBER 2
123
quality of care at the institutional level and conduct quality assessment research. l6 The Patterns of Care Study for Radiation Oncology is the first nationwide evaluation of a clinical discipline and provides a model for other oncology specialties. Department of Urology State University of New York at Buffalo 139 Parker Hall Buffalo, New York 14214 (DR. MURPHY) References 1. Murphy GP: Current status of the National Prostatic Cancer Project treatment protocols, in Denis L, Murphy GP, Prout GR, and Schroder F (Eds): Controlled Clinical ‘IHals in Urologic Oncology, New York, Raven Press, 1984, vol. 13, pp 119-133. 2. Quality Assurance in Radiation Therapy: Clinical and physical aspects. Proceedings of the First Intrnational Symposium on Quality Assessment in Radiation Oncology, Int J Radiat Oncol Biol Phys 10 (suppl. 1): (1984). 3. Gardner A, Bagshaw MA, Page V, and Karzmark CJ: Tumor localization, dosimetry, simulation and treatment procedures in radiotherapy: the isocentric technique, AJR 114: 163 (1972). 4. Hafermann MD: Cancer of the prostate-external radiotherapy, in Murphy GP (Ed): Clinics in Oncology, London,
124
WB Saunders Company Ltd, 1983, vol. 212, pp 371-405. 5. Huben Rp, and Murphy GP: Prostate cancer: an update, CA 36: 274 (1986). 6. Clicksman AS, Reinstein LE, and Laurie F: Quality assurance of radiotherapy in clinical trials, Cancer Treat Rep 69: 1199 (1985). 7. Cunningham JR: Quality assurance in dosimetry and treatment planning, Int J Radiat Oncol Biol Phys lO(supp1. 1): 105 (1984). 8. Glicksman AS, Reinstein LE, MC Shan D, and Laurie F: Radiotherapy quality assurance program in a cooperative group, ibid 7: 1561 (1981). 9. Perez CA, Gardner P, and Glasgow GP: Radiotherapy quality assurance in clinical trials, ibid lO(supp1. 1): 119 (1984). ibid 10. Davis LW: Quality assurance in clinical trials, lO(supp1. 1): 147 (1984). 11. Reinstein LE, Peachey BS, Laurie BS, and Glicksman AS: Impact of a dosimetry review program on radiotherapy in group trials, ibid 11:1179 (1985). 12. Hanks GE, and Kramer S: Consensus of best current management: the starting point for clinical quality assessment, ibid lO(supp1. 1): 87 (1984). 13. Kramer S: The study of the Patterns of Care in Radiation Therapy, Cancer 39: 780 (1977). 14. Kramer S, and Herring D: The Patterns of Care Study: a nationwide evaluation of the practice of radiation therapy in cancer management, Int J Radiat Oncol Biol Phys 1: 1231 (1976). 15. Kramer S: The patterns of clinical care in radiation therapy in the United States, ibid lO(supp1. 1): 49 (1984). 16. Hanks GE: Future plans for quality assurance in radiation oncology in the United States, ibid lO(supp1. 1): 35 (1984).
UROLOGY
/ FEBRUARY
1988
/ VOLUME
XXKI, NUMBER
2