Purpose/Objective: The Patterns of Care Survey process evaluation has been an effective means of assessing the evaluation and treatment provided by Radiation Oncology (RO) practices in the U.S. for Stage I/II breast cancer. The current 1998-9 report updates the previous 1993-4 analysis and reflects recent changes in surgery and systemic therapy observed nationally in the management of early stage disease.
Materials/Methods: 289 patient records of women treated with BCS and RT were randomly selected using a stratified two-stage sampling design and centrally reviewed from RO academic and private practices across the country. Data collected included patient characteristics, clinical and pathologic factors, and surgical and RT treatment details. The results were compared with 1993-4 results using a chi-square test.
Results: Median age was 57 years (range 27-86), similar to the 1993-4 report. Following final surgery, microscopic margins were positive, negative, or close in 9.3%, 68.5%, and 16.6% (unknown=5.5%), respectively, compared to 2.9%, 84.9%, and 7.8% (unknown=4.4%) in 1993-4 (p<0.0001). The percent of primary tumor type was similar to 1993-4 report, but quantification of an intraductal component increased from 8.5% in 1993-4 to 19.4%, p<0.0001. Tumor characteristics were better defined, with estrogen (ER) and progesterone receptors (PR) performed in 83.4% and 82.4% vs. 73.9% and 70.0% in the 1993-4 survey (p=0.001 and p<0.0001, respectively). Fewer specimens were analyzed by flow cytometry in the present analysis (29.1% vs. 49.4%, p<0.0001), but Her-2-neu status was assessed in 32.9%. Sentinel node biopsy (SNB) was performed in 66 patients (22.8%), with positive sentinel nodes (SN) in 12 cases (18%). In patients with positive SN, 92% went to axillary dissection (ALND) vs. 65% if SN were negative (p=ns). 82.4% of all patients had ALND, median 14 nodes removed (range 1-42). As previously observed, 100% received whole breast (WB) RT. Use of CT scans for planning was increased in 1998-9, with 16.6% cases CT-planned vs. 5% in 1993-4 (p<0.0001). A boost was used in 84.4% in 1998-9 vs. 83.8% in 1993-4 (p=ns) using electrons in 96.7%, photons in 2.5%, and mixed beam in 0.8%. The boost was planned on the simulator in 55.7% of patients but clips were present in only 18.7%. A supraclavicular field (SCV) was added in 42% of cases with 1-3 positive nodes and in 88% of cases with 4 or more positive axillary (AX) nodes (p=0.02); full AX RT was only used in 12% and 25% of cases, respectively (p=ns). Similar patterns of regional RT were observed with positive SN followed by ALND, where 45% received SCV RT and 9% full AX RT. Chemotherapy and tamoxifen use was significantly increased compared to 1993-4, i.e., 38.1% and 50.5%, respectively vs. 26.3% and 43.2% in 1993-4 (p=0.0002 and P=0.03) despite comparable pathologic tumor size and nodal involvement.
Conclusions: An increased incidence of close and positive margins (factors known to adversely impact rates of local control) was observed compared to 1993-4. The predictive markers of ER and PR (and Her-2-neu) were more frequently assessed in the current report vs. 1993-4. SNB, not previously available, was utilized in almost 25% of patients in 1998-9. There were no changes in ALND and RT field selection based upon SN findings. Future comparisons will be needed to assess evolving trends. Similar percentages of patients received WB and boost RT as in 1993-4. The estimates in the current survey will serve as a benchmark for future comparisons given the benefit in local control seen using a boost in randomized studies and increasing interest in brachytherapy only to the tumor bed. CT-based planning was more common in the 1993-4 analysis, however better localization of the tumor bed is needed in cases clinically planned. Finally, systemic therapy use continues to increase. Sequencing results will be presented. Supported by NCI grant #CA65435.
Label #0158: White, J., Moughan, J., Pierce, L., et. al.
Purpose/Objective: The indications for PMR are evolving and have been the subject of multiple recent guidelines and consensus statements. The population of breast cancer patients who have actually received PMR has been less well described. The PCS used a two-stage stratified random sample to perform this first known survey to document the practice of PMR in the US for breast cancer patients diagnosed in 1998-99.
Materials/Methods: Data was collected by a PCS research associate from 47 randomly selected institutions (38% academic) who agreed to participate. Patients were excluded from the survey if they had: bilateral breast cancer, prior or concurrent malignancies, previous radiation therapy, or neoadjuvant chemotherapy. A multivariate analysis was performed using tumor stage, < or > 3 positive nodes, and the presence or absence of each of the following: extra-capsular extension (ECE), lymphatic/vascular space invasion (LVI), dermal LVI, and positive surgical margins to determine if they influenced the radiation fields used.
Results: The survey yielded 321 patients: mean age was 54 years (range 24-87), 83% were white and 64.5% post-menopausal. The mean tumor size was 3.74 cm., with 25% T1, 49% T2 and 25% T3. LVI in the breast was noted in 42% of cases, dermal LVI in16%, and margins were positive in 14.6% or close (<2mm) in 17.6%. The mean number of positive nodes was 5.45 (median 3.0) and 32.2% of patients had 1-3 positive nodes, 47.1% had > 4 positive nodes and 20% were node negative. ECE was identified in 27%. Breast reconstruction was performed for 85 patients (26%) with 75 having reconstruction prior to RT (autologous 40, implant 23, both 9, other 3). The mean time to treatment from date of mastectomy to initiation of radiation was 181.4 days (median 188, range 23–489 days). Chemotherapy was delivered pre-radiation for 93%.
Radiation was delivered to the chest wall in 97% of cases and 82% received chest wall and nodal radiation. Photons were much more commonly used for treatment of the chest wall then electrons (87% vs.12%). The mean chest wall dose was 50.2 Gy with 93% receiving between 45 and 50.4 Gy. A chest wall boost was delivered to 44.5% with mean electron energy of 7.8 MeV and a mean dose of 14.8 Gy. The 75 reconstructed breasts underwent radiation to similar doses almost exclusively with photons (97%) and 28% were boosted. Of those who had nodal radiation, a supraclavicular field was delivered in 97%, most commonly with 6 MV photons (83%) to a mean prescribed dose of 48.5 Gy delivered to a depth for 75%. A posterior axillary field was documented in 35%, given with 6 MV photons for 83% to a mean prescribed dose of 36.7 Gy, and most frequently recorded at mid-plane (61%). Internal mammary fields were documented in 59 (22.5%) patients and delivered using a mean energy of 11.37 MeV electrons. Dosimetry was based on a central plane contour for 67% and 3-dimensional or multiple axial planes for 22%. A CT was used for planning in 27% of cases.
Patients were more likely to receive chest wall without nodal radiation when there were 3 nodes positive (odds ratio (OR) 13.4, p<0.0001), and no ECE (OR 5.3, P=0.007). None of the variables analyzed predicted for the use of a chest wall/ scar boost. Only the > 3 positive nodes variable was associated with the use of either a posterior axillary (OR 2.8, p<0.0001) or an internal mammary (OR 2.1, P=0.013) field.
Conclusions: This PCS survey demonstrated that breast cancer patients who were diagnosed in 1998-99 and received PMR had a high proportion of factors associated with an increased risk for local regional failure. PMR typically consisted of a chest wall and supraclavicular field treated with 6 MV photons; the utilization of posterior axillary and internal mammary fields was uncommon. The presence/absence of > 3 positive nodes seems to be the only factor that consistently influenced the treatment fields used. Individualized CT based three-dimensional radiation planning is less commonly employed for breast cancer than other major disease sites.
Supported by NCI grant #CA65435.
Label #0117: Shank, B., Moughan, J., Owen, J., et. al.
Purpose/Objective: To determine the patterns of evaluation and treatment in the U.S. of women with early breast cancer treated with breast-conserving surgery and definitive irradiation in 1993-94, and to compare these with the 1992 ACR Standard for Breast Conservation Treatment.
Materials and Methods: In 1995-96, records of patients treated from 1993-94 at 63 facilities were reviewed by one of two surveyors. At each site, 9-12 charts were randomly selected. Facilities were from 3 strata representative of the distribution of types of practice sites (academic, research-participating, and non-research participating). Of the records reviewed, 727 were eligible for analysis for the Process Survey (PS).
Results: Compared with the last PS in 1983, patients in the 1993-94 study had an older age distribution. In the current study, 70% of patients were > 50 y old, and 69% were post-menopausal, compared with 59% > 50 y old and 49% post-menopausal in 1983, differences which were statistically significant (p=0.0087 and <0.001, respectively).
Workup and evaluation in 1993-94 was closely aligned with the ACR Standard and considerably improved when compared with 1983 (N/E = not evaluated):
Only 3% of patients in 1993-94 were on any typed of clinical trials. In spite of this, radiation treatment parameters closely adhered to ACR Standard recommendations, improving substantially from the 1983 PS:
Conclusion: There was an extensive shift to adherence to the 1992 ACR Standard in 1993-94, when compared with the 1983 PS, in both evaluation and treatment of patients with early breast cancer.
Label #0135: Brenin, D., Morrow, M., Moughan, J., et. al.
Routine axillary lymph node dissection (ALND) for breast cancer patients has become controversial. Factors influencing the performance rates of ALND and axillary irradiation (AI) were evaluated in a joint study of the American College of Surgeons and the American College of Radiology. 17,931 patients with Stage I and II breast cancer treated at 827 institutions in 1994 were studied. 15,992 (93.2%) underwent ALND. The mean ages of patients who did and did not undergo ALND were 60.4 yrs and 73.0 yrs (p<0.0001). Patients with T1a tumors underwent fewer ALND's when compared to patients with larger tumors (81% vs. 93%, p<0.0001). Patients with favorable histology (tubular, papillary and mucinous carcinomas) underwent ALND in 87.9% of cases, compared to 93.6% of patients with other histologies (p<0.0001). Women age 70 or older underwent fewer ALND's compared to younger women (86% vs. 97%, p<0.0001).
ALND rate did not vary between palpable vs. non-palpable tumors nor with tumor grade. 899 patients received AI. Patients undergoing ALND were more likely to receive AI (27% vs 12%, p<0.0001). In patients who underwent ALND, 1.6% of those with no lymph node metastasis received AI, 8.9% of those with 1-3 nodal metastases received AI, 24.0% of those with 4-9, and 29.9% of patients with > 10 nodal metastases received AI.
We conclude that the majority of patients with small breast cancers continue to undergo axillary dissection while AI is under-utilized in patients at risk for local regional relapse.
Label #0134: Morrow, M., Winchester, D., Chmiel, J., et. al.
Guidelines for BCT were developed in 1992 and widely disseminated. This study conducted by
the American College of Surgeons and the American College of Radiology was undertaken to
determine current patterns of care and to evaluate guideline adherence. 17,931 patients with Stage I and II breast cancer treated at 827 institutions in 1994 were studied.
Only 7,914 (44.1%) had BCT. 46.7% of BCT patients were under age 60 compared to 40.7% of mastectomy patients (p<0.0001). Significant differences in clinical and pathologic stage were noted between patients undergoing BCT and mastectomy, with 53.6% of 8,312 clinical stage I patients having BCT compared to 32.2% of 4,138 clinical stage II patients and 38.6% of 5,252 patients with no clinical stage data (p<0.0001). Significant differences in procedure were noted on the basis of both clinical tumor size and nodal status with 52.7% of 9,140 T1 tumors having BCT versus 32.6% of 3,954 T2 tumors (p<0.0001), and 47.2% of 11,435. No patients versus 31.9% of 920 N+ patients having BCT (p<0.0001). These differences persisted when pathologic stage was considered, with 51.4% of 9,662 pathologic stage I patients having BCT compared to 30.5% of 7,417 pathologic stage II patients (p<0.0001). Patients with favorable histologies (tubular, mucinous, intracystic, n=840) were more likely to undergo BCT than those with other histologies (n=17,062; p<0.0001). Radiotherapy (RT) was given to 78.6% of BCT patients. Of 1,155 patients not receiving RT, surgical failure to refer for RT accounted for 51.1%, and patient refusal for 15%. These results indicate that surgeons continue to utilize BCT primarily for patients
with favorable breast cancer, in spite of guidelines and data from randomized trials indicating that age, prognosis, and tumor type should not be used as selection criteria for local therapy. This misunderstanding is a major factor responsible for low national rates of BCT.
Label #0100: Wilson, J.F., Winchester, D., Owen, J., et. al.
A national collaborative study was conducted by the American College of Radiology and
American College of Surgeons during 1995-96. Primary study aims were to 1) determine the fraction of eligible patients treated with breast conservation therapy (BCT) vs. mastectomy and the factors which influence the treatment decision, including practice setting, age and minority status; 2) compare the actual processes of care with the recommendations in the interdisciplinary Standards for Breast Conservation (SBC) (1992); and 3) compare outcomes of BCT vs. mastectomy regarding local and distant failure rates and treatment-related morbidity.
Data was collected on 17,931 Stage I and II breast cancer cases treated during 1994 as reported by over 800 institutions responding through the ACOS Patient Care Evaluation program to a questionnaire which was developed based upon the SBC.
Mastectomy vs. lumpectomy rates among the patients studied were 55.9 and 44.1%, respectively. There was an obvious trend of the use of mastectomy with increasing age with the exception of very young or very old patients. Overall, 15% of patients having lumpectomy as primary surgery did not receive radiation therapy, while 10% of mastectomy patients received adjunctive radiation. An analysis of the variations in practice setting, minority status, clinical presentation, and tumor size between the two treatment groups will be discussed.
This unique interdisciplinary study permits assessment of the current care processes of early breast cancer in the U.S. and an opportunity to refine the SBC where necessary and determines where additional educational efforts are needed.
Label #0109: Coia, L., Wizenberg, L., Gunderson, D., et. al.
Purpose/Objective: Clinical trials of surgical adjuvant treatment in rectal cancer (RC) indicate that radiation therapy with concurrent chemotherapy (CRT) is superior to radiation alone (RT). Whether preoperative treatment is superior to postoperative treatment is controversial. This Patterns of Care Study (PCS) survey patients with RC treated with radiation in 1988-1989 to determine the national practice standards and outcomes and to compare the results of clinical trials.
Materials and Methods: A national survey of 73 institutions using two-stage cluster sampling was conducted and specific information on 406 patients with RC who received radiation from 69 facilities was collected. Follow-up ranged from 0-8.44 years with a median of 4 years. Median age was 64 years. 62% were male, 38% female. 46% underwent low anterior resection and 43% abdominal perineal resection. 195 patients (75%) received postoperative treatment and 65 (25%) received preoperative treatment. The median total radiation dose was 49.05 Gy. 133 patients (43%) received chemotherapy. 5-FU alone was most commonly used (69%) and was commonly given via infusion (62%).
Results: Survival was stage dependent (87.5% stage I, 62.4% stage II, 53.7% stage III at 5 years, p=0.066). Local or regional failure was similar for stage II and stage III (12.8% and 11.4% at 5 years respectively). There was no difference in survival whether patients received preoperative or postoperative treatment (61% vs. 59% at 5 years respectively). Similarly there was no difference in local-regional failure at 5 years (12% pre-op vs. 5% post-op, p=22). A significant difference in survival was seen with post-up CRT vs. Post-op RT (70% vs. 46% at 5 years, p=0.005). No differences in survival or local-regional failure were seen when radiation treatment was interrupted > 3 days. Multivariate analysis indicated the only treatment related variable of prognostic importance for survival was the use of CRT.
Conclusion: This study confirms the value of adjuvant post-op CRT over post-op RT in RC treatment. No differences in survival or local-regional failure were seen with pre-op vs. post-op adjuvant treatment. As treatment was not determined by randomization, these results could be a function of selection.
Label #0104: Minsky, B.D., Coia, L.R., Stouffer, N.O., et. al.
Purpose:To determine the U.S. national practice for the treatment of adenocarcinoma of the rectum.
Materials and Methods:From a national survey sample of 49 institutions (15: academic, 14: research participating, 20: non-research participating) 460 patients who received radiation therapy (RT) as a component of their treatment for rectal cancer were sampled. A stratified two-stage cluster sampling with simple random sampling at each stage for each stratum was used and on-site surveys were performed by 2 Research Associates.
Results:Of the 460 pts., 347 (75%) received post-operative therapy, 97 (21%) had pre-operative therapy, 14 (3%) had pre + post-operative therapy, and 2 had intraoperative RT alone. A local excision was performed in 12% of pts. compared with 3% in the 1988-1989 PCS process survey. Overall, 84% of pts. had a medical oncology consultation and a total of 80% received chemotherapy compared with 44% in the 1988-1989 PCS process survey. In order to assess the penetration of modern RT techniques as well as recommendations of Intergoup clinical trials into current practice this analysis is limited to the 101 patients (22%) who underwent conventional surgery (low anterior resection or abdominoperineal resection [APR]) with negative margins and had T3 and/or N1-3M0 disease. Since 1991, the standard of care for this group of pts. has been post-operative combined modality therapy consisting of pelvic RT + 6 months of chemotherapy. Although only 7% of the 101 pts. were treated on a clinical trial (6%: NCI Intergroup and 1%: institutional), 92% received chemotherapy for a median of 6 months (range: 0.5-28.5 months). Likewise, most were treated with modern RT treatment techniques (68%: prescribed to an isodose line, 80%: ˇ 10 MV protons, 92%: 3-4 field technique, 84% prone, 82%: all fields treated each day, and 93%: APR scar treated in all fields. Median doses included 1.8 Gy/fraction, 45 Gy to the pelvis (range: 16.2-54 Gy), plus a 8.5 Gy boost (range: 3.6-16.0 Gy), delivered over a total median elapsed time of 42 days (range: 16-85 days). However, other standard procedures were not routinely used either at the time of surgery (53%: no clips placed, and 54%: no attempt to exclude small bowel) or during the RT simulation (46%: no small bowel contrast, and in patients who did not undergo an APR, 32% had no rectal contrast).
Conclusions:Despite that only 7% of eligible patients were placed on a clinical trial, these trials have had a positive influence on the standard of practice within the oncology community. Although there are still some deficiencies, the majority of patients with stage T3 and/or N1-3M0 disease receive combined modality therapy and are treated with modern RT techniques. Educational efforts need to be continued in the areas of small bowel identification and exclusion.
Label #0165: Suntharalingam, M., Moughan, J., Coia, L., et. al.
Purpose/Objective: A Patterns of Care study (PCS) was recently conducted in order to evaluate the standards of practice for patients receiving radiation therapy for esophageal cancer from 1996-1999. This study examined the evaluation and treatment schemes utilized during this time period, and compared these results to the original PCS data obtained between 1992-1994 in order to identify any fundamental changes in national practice.
Materials/Methods: A national survey was conducted using 2- stage cluster sampling technique. Specific information was collected on 334 pts with esophageal cancer who received radiotherapy (RT) as part of definitive or adjuvant management at 47 institutions. Patients were staged according to both the AJCC 1983 and 1997 UICC systems. Eligibility criteria for case review included, RT between 1996-1999, no evidence of distant metastases (including CT evidence of either supraclavicular or celiac nodes > 1 cm), squamous cell or adenocarcinoma histology, KPS ³ 60, tumors in the thoracic esophagus with < 2 cm extension into the stomach, and no prior malignacies within the last 5 years.
Results: The median age of pts was 65 years, 77.8% were males and 22.2% were female. KPS was > 80 in 81.7% of patients. The racial profile mirrors the previous study with 76% Caucasian, 19% African American, 3% Asian and 1.2% Hispanic. Squamous cell carcinoma was diagnosed in 52% of patients and adenocarcinoma in 47%. 13% were clinical stage (CS) I, 36% CS II, and 37 % CS III according to the '83 AJCC system. Work up included endoscopy (96%), CT of the chest (87%), CT of the abdomen (74%), and esophagram (62%). Endoscopic ultrasound (EUS) was used in 19% of cases as compared to 3.5 % from the previous survey (p < 0.0001). Patients treated at an academic center were more likely to undergo EUS than those treated in a non-academic setting (29% vs. 10%, p =0.0001). Fifty-nine percent of patients received concurrent chemoradiation as definitive treatment. There was a significant increase in the use of concurrent chemoradiation prior to planned surgical resection as compared to the original study (21.3% vs. 13.4%, p= 0.004). Other schemes included RT alone (12%), post-op RT (2%), and post op chemoradiation (5%). Forty-three percent of patients with adenocarcinoma underwent tri-modality therapy, as compared to 13% of those with squamous cell carcinoma (p<0.0001). Patients undergoing tri-modality therapy were more likely to have had an EUS (p=0.007).
The median total dose of external RT was 50.4 Gy, and median dose per fraction was 1.8 Gy. Brachytherapy was used in 8% of cases. The chemotherapy agents most commonly utilized included 5-FU (84%), Cisplatin (66%), and paclitaxel (20%). Paclitaxel was more commonly employed as part of a preoperative chemoradiation regimen than in the setting of definitive chemoradiation (37% vs. 12% p <0.0001). Compared to the '94 survey, paclitaxel use significantly increased between '96-99 (0.7% vs. 20%, p<0.001).
Conclusions:This Patterns of Care Survey confirms the use of concurrent chemoradiation as part of the national standards of practice for management of esophageal cancer patients. A comparison with the previous study documents the significant rise in the use of EUS, tri-modality therapy, and the increasing utilization of paclitaxel as part of a combined modality regimen.
Label #0147: Teshima, T., Owen, J., Hanks, G., et. al.
Purpose: To improve the quality of Japanese radiation oncology, the Patterns of Care Study (PCS) has been imported from the United States to Japan. An international comparison of PCS results for esophageal cancer patients in the US and Japan was performed with special reference to age to determine the current process for the elderly ($75 years old) in the two countries, because this population has not been covered by most previous clinical studies.
Materials and Methods: The same data format as that used for the PCS in the US was provided by courtesy of the American College of Radiology (ACR). The Japanese study was supported by the Ministry of Health and Welfare (8-27 and 29). From July 1996 to February 1997, the process for radiation therapy for 455 esophageal cancer patients in 29 institutions nationwide in Japan (approximately 5 %of all institutions) were surveyed by means of extramural audits. Corresponding data for the US were collected for 400 esophageal cancer patients by original two-staged cluster sampling from April 1995 to July 1997. The process for these patients in both countries were compared for two age groups: those aged 75 years or older (elderly group; US: 80 patients, Japan: 113) and those aged 74 years or younger (younger group; US: 320, Japan: 342).
Results: There were significant differences in the types of institutions where patients were treated, in their background, and in the treatment for the two age groups in both countries. As for type of institution (Table), more elderly patients were treated in non-academic than in academic institutions in both countries (US: p5.0044, Japan: p5.0050, US vs Japan $75 y/o only: p5.1584). The proportion of female patients in the elderly group was significantly higher by 10% or more than that in the younger group in both countries (US: p5.0338, Japan: p,.0001, US vs Japan $ 75 y/o only: p5.2391). As for treatment, the ratio of surgery for the elderly group was only one-third of that for the younger group (US: p5.0005, Japan: p,.0001, US vs Japan $75 y/o only: p5.2796). The application of chemotherapy for the elderly group was less by 20% or more than that for the younger group (US: p5.0002, Japan : p,.0001, US vs Japan $75 y/o only: p,.0001). For non-surgery patients in both countries, there was no significant difference in the dose range between the two age groups (US: p5 .3800, Japan: p5.3001, US vs Japan $75y/o only: p,.0001).
Conclusion: 1) Significant difference in types of institutions, patients' background and treatment for esophageal cancer between the two age groups were observed in both countries. 2) More elderly patients were treated in non-academic than in academic institutions in both countries. 3) There were significant decreases in the use of surgery and chemotherapy for the elderly group. 4) Radiation therapy was used with equal dose ranges for both age groups. 5) This information should be useful for future prospective studies of the elderly with radiation therapy.
Label #0139: Teshima, T., Owen, J., Hanks, G., et. al.
Purpose: To improve Japanese quality of radiation oncology, the Patterns of Care Study (PCS) has been imported from the United States to Japan. International comparison of the PCS for esophageal cancer patients in the US and Japan was performed to determine current quality of care in two countries.
Materials and Methods: The same data format as that used for PCS in the US was provided by courtesy of American College of Radiology (ACR). The Japanese study was supported by the Ministry of Health and Welfare (8-27 and 8-29). From July 1996 to February 1997, the process of radiation therapy for 455 esophageal cancer patients in 29 institutions nationwide in Japan (approximately 5% of all institutions) were surveyed by means of extramural audits. Corresponding data for the US were collected for 399 esophageal cancer patients by original double randomized sampling from April 1995 to July 1997. The processes for these patients in both countries were compared.
Results:There were significant differences in patient presentation and treatment in the US and Japan. For the background (Table 1), 36% of patients in the US had adenocarcinoma, but in Japan, 98% had squamous cell carcinoma (p<.0001). In the US, 31% of patients were classified as Stage III by AJCC, while 53% in Japan (p<.0001). For the treatment (Table 2), surgery was performed on 26% of patients in the US, but in Japan, 40% received surgery (p<.0001), while 77% of patients in the US received chemotherapy and 42% in Japan (p<.0001). For non-surgery patients in the US, moderate to high doses from 40 to 64 Gy were delivered to 83% of patients, but in Japan, 73% of patients were irradiated with higher doses from 60 to 74 Gy. There was thus a significant difference in the dose range between the two nations (p<.0001).
Conclusions:1) Significant difference in patient presentation and treatment for esophageal cancer in the US and Japan were observed. 2) Higher utilization of chemoradiotherapy for esophageal cancer in the US was suggested by a shift of the external dose range to moderate and the higher incidence of combined chemotherapy than that used in Japan. 3) Surgery was more frequently performed in Japan for esophageal cancer than in the US. 4) Further follow-up for these patients is needed to investigate the influence of such significant differences on outcomes. 5) Other factors such as different medical care systems should also be taken into account in order to obtain a broader view.
Label #0132: Coia, L.R., Minsky, B.D., John, M., et. al.
Background and Objective:For the first time, a Patterns of Care Study (PCS) was conducted in 1992 - 1994 to determine the national practice standards in evaluating and treating esophageal cancer and to test penetration of clinical trials into national practice.
Methods:A national survey of 61 institutions using two-stage cluster sampling was conducted and specific-information on 400 patients with squamous cell or adenocarcinoma of thoracic esophagus who received radiation (RT) as part of definitive or adjuvant management was collected. Patients were staged by a modified 1983 AJC staging system. Tests for significant differences between academic and non-academic institutions for a particular variable were performed using a chi-square statistic.
Results:The median age was 66.7 years (range 26-89 years), 76.5% were male, and 23.5% were female. KPS was > 80 in 88.3%. Squamous cell cancer was diagnosed in 61.5% and adenocarcinoma in 36.8%. 15% were Clinical Stage (CS) I, 39.5% CS II, 29.5% CS III. Evaluative procedures included endoscopy (>93%), CT of the chest (86%), CT abdomen (75%), esophagram (68.5%) and endoscopic ultrasound (3.5%). Endoscopic ultrasound and CT of the chest were performed significantly more frequently in academic than non-academic facilities (6.1% vs. 1.0% and 91.9% vs. 81.3% respectively). Three-quarters of all patients received chemotherapy and RT and 62.5% received concurrent chemo and RT as part of their treatment. Treatments included chemo plus RT (54.0%), RT alone (20.3%), pre-op chemo + RT (13.3%), post-op chemo + RT (7.7%), post-op RT (3.5%) and pre-op RT (1.2%). Chemo agents most frequently used were 5-FU (84%), cisplatin (64%) and mitomycin (9%) with academic institutions using cisplatin significantly more and mitomycin significantly less than non-academic institutions. Brachytherapy was used in 8.5%. The median total dose of external beam radiation was 50.4 Gy and median dose per fraction was 1.8 Gy.
Conclusions:This study establishes the national benchmarks in the evaluations and treatment of esophageal cancer in radiation facilities in the U.S. It also indicates that the majority of patients treated with radiation as a component of treatment for esophageal cancer received chemoradiation rather than RT alone as supported by clinical trials. Although some differences in evaluation of esophageal cancer were noted between academic and non-academic facilities, there was no differences in the frequency of use of chemoradiation versus radiation by facility type.
Label #0133: Coia, L., Minsky, B., John, M., et. al.
Clinical trials in esophageal cancer (EC) show that radiation therapy (RT) with concurrent chemotherapy (CRT) is
superior to radiation alone and that preoperative CRT may be superior to esophagectomy. A Patterns of Care Study (PCS) examined the records of patients with EC treated with radiation in 1992-1994 to determine the national practice standards and outcomes and to compare the results to clinical trials. A national survey of 61 institutions using two-stage cluster sampling was conducted and specific information on 400 patients with squamous cell or adenocarcinoma of the thoracic esophagus who received RT as part of primary or adjuvant treatment was collected. Patients were staged by a modified 1983 AJC staging system. Squamous cell cancer was diagnosed in 61.5% and adenocarcinoma in 36.8%. 15% were clinical stage (CS) I, 39.5% CSII, 29.5% CSIII. Median total RT dose was 50.4 Gy. 26% of patients underwent esophagectomy. 75% of patients received chemotherapy; 83% of these received CRT. Chemotherapy agents most frequently used were 5FU (84%), cisplatin (64%) and mitomycin (9%). Follow-up ranged from 0-5 years with a median follow-up of 9 months. Significant variables for overall survival in multivariate
analysis include the use of esophagectomy (RR=.64), the use of chemotherapy (RR=.64), KPS <80 (RR=1.56) and clinical stage III (RR=1.49). Age, sex and histology were not significant. Preoperative CRT results in significantly higher survival (p=0.04) and a trend toward higher local control (p=0.09) than CRT alone; however, these differences did not exist under multivariate analysis. This study confirms the value of CRT in EC treatment and indicates the usefulness of KPS and the 1983 clinical staging system as prognostic indicators. The value of esophagectomy and the suggested superiority of preoperative CRT over CRT should be tested in a randomized trial.
Label #0160: Zelefsky, M., Moughan, J., Owen, J., et. al.
Purpose/Objective:To report changing trends in external beam radiotherapy (ERT) delivery practice for clinically localized prostate cancer based on the 1999 survey from the American College of Radiology National Patterns of Care Study (PCS).
Materials/Methods:The 1999 survey included 443 patients surveyed from 47 institutions treated between 1/1999 and 12/1999. Of these, 123 (28%) were treated with brachytherapy + ERT and 320 (72%) were treated with ERT only; this latter group are the subjects of this report. For this analysis, patients were categorized into the following risk groups: favorable-absence of adverse features (PSA<10, Gleason Score <6 and T stage < T3); the presence of one or more of these features classified patients into intermediate and unfavorable risk groups, respectively. 40%, 39% and 21% were categorized as favorable, intermediate and unfavorable groups, respectively. The following trends in clinical practice were analyzed according to these prognostic risk groupings and other variables and compared to the prior survey from 1994: the use of androgen deprivation therapy in combination with ERT, higher prescription dose levels, and the administration of elective whole pelvic radiation (WPR).
Results:The median age in this group was 72 years (range: 49–86 years). The clinical stages in these patients were: TX-6%, T0–6%, T1–43%, T2–34%, T3–6%, T4–0.3% and unknown in 6%. The pre-treatment PSA levels were: (<4)-11%, (4–10)-57%, (11–20)-20%, and (>20)-12%. The Gleason score breakdown was (2–6)-53%, (7)-29% and (8–10) in 18%. 160 (50%) patients were treated with androgen deprivation therapy (ADT) in conjunction with ERT. Conformal radiotherapy was reported to be used in 85% of treated patients. The median prescription dose was 70.45 Gy (range: 27–79.2 Gy). ADT usage was significantly more prevalent among treated patients in the 1999 survey compared to the 1994 survey (50% versus 9%, p < 0.0001). In the 1999 survey the incidence of ADT usage for favorable, intermediate and unfavorable risk groups were 30%, 54% and 81%, respectively. In a multivariate analysis PSA >10 and Gleason scores of 7–10 were the only significant variables influencing ADT usage in the 1999 survey. Compared to the prior survey, a greater percentage were treated with higher radiation doses (>72 Gy, 1999–48% vs. 1994–3%, p < 0.0001). In the 1999 survey the percentage of patients with favorable, intermediate and unfavorable tumors treated to doses >72 Gy were 48%, 43% and 56%, respectively, compared to 3%, 3% and 4%, respectively, in the 1994 survey. In a multivariate analysis > T3 stage and academic centers versus community-based centers were the only significant variables influencing high-dose ERT delivery in the 1999 survey. Compared to the 1994 survey, there was a significant increase in brachytherapy (28% vs. 3% p < 0.001). There was no significant difference observed for the use of WPR in the 1999 compared to the 1994 survey (1999–28% vs. 1994–33%, P = 0.17). For favorable, intermediate and unfavorable risk patients in the 1999 survey, WPR was used in 16%, 32% and 44% respectively. Intermediate/unfavorable risk patients, regions (North Central/South), and > T3 stage were the only significant variables influencing the use of WPR delivery.
Conclusions:The significantly increased usage of ADT for high-risk patients and higher radiation doses especially for intermediate and high-risk patients reflect the penetrance and growing acceptance of clinical trial results which demonstrate the efficacy of these treatment approaches. The relatively high percentage of favorable risk patients treated with high doses is greater than expected. A significant increase in brachytherapy was observed compared to prior surveys. The majority of treated patients with high-risk disease do not receive elective WPR, which likely reflect influences of prior trials, stage migration and increased use of conformal radiotherapy treatment planning.
Label #0141: Chuba, P.J., Moughan, J., Owen, J., et. al.
Purpose/Objective:To analyze the five year results from the 1989 patterns of care (PCS) study for prostate cancer. This is the first PCS to have collected pre and post treatment serum PSA data and to have included large numbers of patients treated with conformal methods.
Material and Methods:600 prostate cancer patients treated with curative radiotherapy at 71 institutions in 1989 made up the study population. Of the 389 analyzable cases, 242 had pretreatment PSA values obtained and 243 had a Gleason's sum (GS) reported. 356 were Caucasian, 24 African American and 3 Hispanic (6 unknown). 102 patients had PSA <10, 60 had PSA 10-19, and 92 presented with PSA greater than 20. 97 patients were from RTOG, CCC or teaching institutions, 141 patients were from other hospital based, non-teaching institutions, and 151 were from free-standing radiation oncology facilities. 70 patients were T1, 202 T2, and 100 T3/4. 24 out of 389 patients also received neoadjuvant hormone therapy. Survival curves were constructed using Kaplan-Meier methods and differences between groups were tested for significance using the log-rank test. A patient was considered a PSA failure if the PSA was > 1.5 and rising.
Results:With a median follow-up of 4.7 years, five year bNED and overall survival were 62% and 76% for stage T1 respectively. The figures for T2 were 58% and 79%, and 39% and 59% for stages T3 and T4 combined. bNED survival was adversely impacted by T stage (p=0.0125), and pretreatment PSA (p=0.001), but not significantly by the GS (GS 2-7 vs. 8-10; p=0.059). Cause specific survival was significantly lower in patients with higher T-stage (p=0.0034), GS (p=0.004), and also pretreatment PSA (p=0.0001). Overall survival was also significantly lower in patients with higher T-stage (p=0.0213) and GS (p=0.023) but not pretreatment PSA however (p=0.111). No associations between clinical local-regional recurrence and T-stage, PSA, or GS were observed. Few late complications were reported - 13/389 and 8/389 grade 2-3 GI and GU complications respectively, with one patient having required a permanent colostomy.
Conclusion:Increased pretreatment PSA was a strong predictor of both biochemical failure and death due to prostate cancer (CSS). A minority of recurrences were detected by clinical exam. New strategies for patients with high stage, high grade tumors and/or pretreatment PSA >20 deserve testing.
Label #0140: Zietman, A., Owen, J., Moughan, J., et. al.
Purpose:To determine the impact of published findings on the contemporary patterns of radiotherapy practice to men with prostate cancer in the US and to evaluate the levels of care received by minority patients.
Materials and Methods:81 institutions were randomly selected for the National Study (NS) from a master list of all US radiation facilities and 63 visited. To study the influence of patient ethnicity on practice patterns 17 additional facilities were selected for a Minority-Rich Survey (MRS). Each facility submitted a list of patients treated in 1994 for apparently localized cancer of the prostate. Patients were randomly selected from each institution for review. A total of 926 patients' records were reviewed in the treating institution by one of two trained research associates (595 fort he NS and 331 for the MRS).
Results:In the NS 6% of men had tumors of palpation stage Tx, 30% T1, 53%, T2, and 9% T3-4. In 90% the diagnosis was made using TRUS biopsy and 3% by TURP. CT scan was part of staging in 85%, MRI in 9%, and lymph node dissection in 5%. 16% had Gleason 2-4 disease, 67% Gleason 5-7, and 15% Gleason 8-10. The median pretreatment PSA was 10.2 ng/ml. External beam as sole radiation modality was given to 97%, brachytherapy to 2%, and a combination to 1%. Nine percent of all patients received adjunctive hormones therapy. External beam was delivered using linear accelerators in virtually every case and only 18% were treated with energies of <6MV (47% >18MV). CT scans were used in planning 79%. The median and modal radiation dose delivered was 68.4 Gy with a modal fraction size of 1.8 Gy. Radiation dose of >70 Gy were given to 23% overall: 19% of T1, 21% of T2, and 36% of T3-4. Doses >70 Gy were given to 17% of those with PSA values of <10, 26% 10-<20, and 30% >20 ng/ml. Thirteen percent of the NS patients were African-American (AA) and 29% in the MRS. In the NS 33% of Caucasian men had T1 tumors, 36% T2a-b, 15% T2c, and 8% T3-4 as compared with 15%, 34%, 28%, and 12% for AA men. In the MRS, however, the proportions were similar between races. In the NS 50% of Caucasian men had PSA values of <10, 28% 10-19, and 18% >20 ng/ml. The corresponding figures were 27%, 29%, and 44% for AA men. Comparable differences were seen in the MRS. Hormone therapy was given to both races with similar frequency (9%) although the proportion of AA men with >T2c tumors was substantially higher.
Conclusions:This study shows several trends since the 1989 study. Men are treated with earlier T stages and with lower median PSA values reflecting the impact of early detection programs. External beam remains the mainstay of radiation treatment though early evidence of interest in brachytherapy is seen. External beam practice is now exclusively linear accelerator and most often CT planned. Median radiation doses are steady though a higher proportion receive doses >70 Gy. Hormone therapy is more frequently used reflecting the integration of newly published data into practice. AA men are less likely to be treated with T1 disease or PSA values of <10 ng/ml. The later held up in the MRS where access to early detection can be presumed to be similar between races. This suggests either that AA men have a more virulent form of the disease or are less responsive to early detection programs.
Label #0142: Katz, A., Eifel, P., Moughan, J., et. al.
Purpose/Objective:To describe the relationship between socioeconomic variables such as race, age and income and factors that affect outcome, including clinical presentation, treatment approach and patient compliance.
Materials and Methods:Sixty-one institutions (NS) were randomly selected for the Patterns of Care survey from a master list of all facilities delivering radiotherapy in the United States, stratified according to practice type. Each institution was asked to provide a list of all patients treated during 1992-1994 with radiation for squamous cell carcinoma of the cervix, from which cases were randomly selected for review. Patients treated with initial hysterectomy or those with distant metastasis were excluded. A total of 471 patients' charts were reviewed for information about demographics, clinical presentation and treatment approach. The study sample was supplemented with 17 minority rich institutions (MRS), which were randomly selected from a list of facilities treating more than 40% African American, Hispanic, Asian/ Pacific Islander or Native American patients. 251 patients' charts were similarly reviewed at these institutions. The median household income of each patient's neighborhood was identified by matching their zip code to 1990 United States Census data.
Results:There was no significant correlation between patients' minority status and their FIGO stage, tumor size, Karnofsky performance status (KPS), or the likelihood of having completed treatment. Patients who were < 40 or > 60 years old at diagnosis tended to present with later stage disease than those aged 40-60 years (p = 0.01). Overall, 85% of the patients surveyed completed the planned treatment. Young patients (< 40 years) in the MRS were less likely to complete treatment than patients who were 40-60 or >60 years of age (p = 0.002). However there was no significant correlation between age and treatment completion in the NS.
In all, 393 of the patients in the NS completed treatment with external beam + low dose rate brachytherapy. Of these, patients belonging to minority groups were more likely to be treated with < 1 implant than non-Hispanic whites (29% vs. 45%, p = 0.006). Overall, patients in MRS institutions were more likely to receive only a single implant; within MRS facilities, minority patients also tended to have a single implant more often than other patients (39% vs. 56%) but the difference was not significant. In both the MRS and the NS, the median total dose to Point A was about 5 Gy higher for patients who had more than one implant compared with those who had only one (84.7 Gy vs. 79.2 Gy and 84.7 Gy vs. 80.0 Gy, respectively).
Of minority patients in the MRS and NS, 45% and 36%, respectively, lived in neighborhoods with a median household income below $20,000. In contrast, only 23% and 14% of non-minority patients in the two surveys lived in these low-income neighborhoods. No significant correlation was found between pretreatment factors (FIGO stage, tumor size, or KPS) and median neighborhood income. Patients in the MRS who lived in higher income neighborhoods were significantly more likely to have > 2 intracavitary procedures than patients from poorer income areas who completed their planned treatment (26% vs. 54%, p = 0.003). In the NS, 19% of patients with Stage I-IIA disease from low income neighborhoods and 23% from higher income neighborhoods underwent adjuvant hysterectomy. In the MRS, fewer patients with early disease had adjuvant surgery; furthermore, only 5% of patients from low income neighborhoods had adjuvant hysterectomy vs. 16% of patients from higher income neighborhoods.
Conclusion:No significant association was found between ethnicity or neighborhood income and the extent of disease at diagnosis. Compliance with planned treatment also was similar for minority patients, although there was some suggestion that young patients, particularly in lower income neighborhoods had more difficulty completing treatment. However, minority patients and those from lower income neighborhoods, particularly those treated in large, minority-rich institutions did tend to receive somewhat different treatment (fewer brachytherapy treatments, lower RT doses, and less frequent adjuvant hysterectomy) than non-minority and higher income patients. These data suggest the need for further, more detailed studies of the relationship between socioeconomic factors, treatment, and outcome to accurately identify possible disparities in the quality of treatment for these patients.
Label #0105: Eifel, P.J., Owen, J.B., Stouffer, N.O. et. al.
Purpose:To determine the impact of evolving technology and the influence of published findings from retrospective and prospective studies on the patterns of radiotherapy practice for patients with carcinoma of the uterine cervix.
Materials and Methods:66 institutions were randomly selected for the National Survey (NS) from a master list of all radiation facilities in the United States, proportionally stratified according to practice type. To study the potential influence of patient ethnicity on practice patterns, 12 additional facilities were selected for a Minority-Rich Survey (MRS) from a list of institutions that reported a high proportion of minority patients in their census. Each facility submitted a list of patients treated in 1992-94 with radiation for squamous carcinoma of the cervix. Patients who had clinical evidence of distant metastases or who were treated with initial hysterectomy were excluded. Patients were randomly selected from each institution for review. A total of 601 patients’ records were reviewed in the treating institution by one of two trained research associates. Information was recorded about patients’ characteristics, diagnostic evaluation, tumor extent, treatment approach, and radiotherapy techniques.
Results:Of 601 patients, 31%, 40%, 24% and 4% had FIGO Stage I, II, III, or IV disease, respectively. 79% of women had a Karnofsky status ˇ 90 and 76% had a pre-treatment Hgb ˇ 10. Black women tended to have lower hemoglobin levels and a poorer performance status than Hispanic or non-Hispanic white patients although the distribution of tumors according to FIGO stage was similar for women of different ethnic groups.
Diagnostic work-up included a CT scan in 411/516 patients for whom information was available (80%). Of these, 61 (15%) were interpreted as showing evidence of nodal metastasis in either the abdomen or pelvis. Lymphangiogram or MRI were obtained in only 4% and 5% of patients, respectively. 104 patients had surgical exploration for nodal evaluation. Of these, 43 (41%) were found to have regional metastases.
Fields were designed using a dedicated simulator in 97% of 563 cases for whom information was available. However fields were designed using only a diagnostic X-ray unit in 17 cases. 108 patients (18%) had CT-based treatment planning using a dedicated unit or CT simulator.
582/601 patients (97%) were treated with curative intent. 532 (91%) were treated with a combination of external beam irradiation and brachytherapy. Most of those treated with external alone had locally advanced disease but 18/235 patients with Stage I-IIA disease (8%) did not receive brachytherapy. Patients in the MRS were less likely to be treated with shaped fields (73% vs. 95%), but were more likely to be treated with daily fractions ¨180 cGy (95% vs. 79%), four-field technique (88% vs. 76%) and high-energy photons ˇ 15 MV (85% vs. 57%) than patients in the NS. 111 patients (18%) received the external beam and brachytherapy components of their treatment in different facilities. Only 20 patients (3%) were treated with interstitial templates. Of 520 patients who had intrcavitary treatments, 467 (90%) were with low-dose rate (LDR) sources, 51 (10%) were with high-dose rate (HDR) sources, and 2 were treated with a combination of dose rates. 50% of patients treated with LDR had 1 application with or without adjuvant hysterectomy and 50% had ˇ 2 applications. A broad variety of schedules were used for HDR treatment. The mean overall duration of treatment for all patients was 57 days. However, for 25% of patients, the overall duration of treatment was > 10 weeks. The median duration of treatment for patients with Stages I-IIA was 55 days vs. 62 days for those with Stages IIB-III. 86 patients (14%) failed to complete planned radiotherapy. Patient compliance was more likely to be cited as the reason for incomplete treatment in the MRS (53% vs. 30%) and these patients were also more likely to have unplanned treatment breaks (11% vs. 5%) than those in the NS. 38 patients (6%) were treated on GOG or RTOG protocols and another 10 (1.7%) were treated on institutional IRB-approved protocols. 134 patients (22%) received chemotherapy before (14 pts), during (123 pts), or after (6 pts) radiotherapy. Cisplatin, 5-FU, and hydroxyurea were the most commonly used drugs.
Conclusions:Most patients with carcinoma of the cervix continued to be treated with a combination of external beam radiotherapy and LDR brachytherapy. Although most patients were treated with appropriate treatment planning and modern equipment, the failure to use brachytherapy and a tendency to excessive treatment prolongation may still have compromised treatment in some cases. Despite a lack of convincing data supporting its use, many patients were treated with chemotherapy outside investigational studies.
Label #0159: Movsas, B., Moughan, J., Komaki, et. al.
Purpose/Objective: For the first time, a PCS-Lung was conducted to determine the national patterns of RT practice in patients (pts) treated for non-metastatic lung cancer in 1998-1999.
Materials/Methods: A national survey of randomly selected RT institutions in the U.S. was conducted using two-stage cluster sampling, stratified by practice type (academic vs non-academic). Pts with AJCC 1997 non-metastatic lung ca (KPS > 60), who received RT as definitive or adjuvant therapy, were randomly selected. A total of 444 pt records from 47 institutions were reviewed by trained research associates.
Results:The histologies were small cell lung cancer (SCLC) in 14% vs non-SCLC (NSCLC) in 86% (squamous 34%, adenoca 26%, large cell 5%, NSCLC unspecified 18%, adenosquam 1%, other 2%). The median age was 67 yrs (range 37-92), 58% male, 42% female; 32% were current smokers & 59% had co-morbidities, mostly cardiopulmonary. KPS was > 80 in 84%. While 98% had chest CTs for staging, bone scans and brain imaging were not obtained in 33% and 52% of clinical stage (CS) III NSCLC patients, respectively. PET scans were utilized in 4%. For NSCLC, 12% were CS I, 16% CS II & 54% CS III (4% unknown). The mediastinum was pathologically staged in 31% of NSCLC.
Treatment (tx) strategies varied significantly by histology and stage (Table), p<0.0001. In SCLC & CS III NSCLC, chemo+RT was utilized more than RT alone (p<0.0001); in CS I NSCLC, RT was the primary tx (p<0.0001). For SCLC, the median RT dose was 51Gy, 80% @ 1.8-2Gy/fraction (fx); 10% received hyperfractionated (BID) RT and 23% prophylactic cranial irradiation (PCI—median dose 30Gy/15 fx). For NSCLC, the median RT dose was 60Gy, 85% @ 1.8-2Gy/fx. 6% had split-course RT, 1% a brachytherapy boost & 8% did not complete the planned RT. Common photon energies were 6MV (44%) and 10MV (21%); cobalt was utilized in only 1 pt. CT-treatment planning was used in 50%. Of these, 52% reported utilizing ''3D conformal RT''. In CS III NSCLC patients, CT-planning was utilized more in pts receiving chemo+RT (59%) vs RT alone (37%), P=0.003. The avg. maximum spinal cord dose was 41Gy, but this was ''unknown'' in 29%.
Overall, 57% received systemic tx. Factors correlating with use of chemo included age (70% < 70 yrs vs 34% > 70 yrs, p<0.0001), histology (90% SCLC vs 51% NSCLC, p<0.0001), increasing CS (p<0.0001), increasing KPS (p<0.0001) and lack of co-morbidities(p=0.0002), but not academic vs non-academic facilities (p=0.81). Of all chemo pts, 73% received it concurrently with RT. Carboplatin & paclitaxel were used in NSCLC > SCLC (p=0.0004) vs the reverse for cisplatin & etoposide (p<0.0001). Systemic tx pre-RT occurred in 48% and post-RT in 22%. Surgery was performed in 28% of NSCLC pts (58% lobectomy, 19% pneumonectomy, 11% wedge, 10% exploratory). 5% of all pts were treated on cooperative group trials and 2% on IRB-approved institutional trials.
Conclusions:This study establishes the general patterns of care for lung ca in RT facilities within the U.S. Further improvements in staging (e.g., for CS III NSCLC) and documentation of spinal cord doses should be encouraged. As supported by clinical trials, pts with limited stage SCLC and CS III NSCLC received chemo+RT > RT alone (p<0.0001). As studies supporting BID RT and PCI in SCLC were published in 1999, the penetration of these trials will be assessed in the next PCS-Lung survey, as well as analyses of outcomes, and the application of functional imaging and 3-D conformal RT. Supported by NCI grant #CA65435.
Label #0172: Movsas, B., Moughan, J., Komaki, R., et. al.
Purpose: For the first time, a Lung Patterns of Care (PCS) study was conducted to determine the national patterns of radiation (RT) practice in patients (pts) treated for non-metastatic lung cancer in 1998-1999.
Materials/Methods:A national survey of randomly selected RT institutions in the U.S. was conducted using two-stage cluster sampling, stratified by practice type. Pts with AJCC 1997 non-metastatic lung ca (KPS >60), who received RT as definitive or adjuvant therapy, were randomly selected. To determine national estimates, sample size was weighted by the relative number institutions/per strata and the number of patient records reviewed per the number of patients eligible (per institution). The weighted sample size of 42,335 patient records from 58 institutions were reviewed by trained research associates. The unweighted sample size (or number of patients) was 541.
Results:The histologies were small cell lung cancer (SCLC) in 14.5% vs non-SCLC (NSCLC) in 85.5%. The median age was 67 yrs (range 29-92); 61% were male and 38% were current smokers. While 98% had chest CTs for staging, bone scans and brain imaging were not obtained in 34% and 52% of clinical stage (CS) III NSCLC patients, respectively. Treatment strategies varied significantly by histology and stage, p<0.0001. For example, in SCLC & CS III NSCLC, chemo + RT was utilized significantly more than RT alone; in CS I NSCLC, RT was the primary tx. CT-treatment planning was used in 49%. Of these, 52% reported utilizing "3-D conformal RT". In CS III NSCLC patients, CT-planning was utilized more in pts receiving chemo + RT (61%) vs RT alone (30%), p=0.047.
Overall, 58% received systemic therapy. On both univariate and multivariate analyses, factors correlating with increased use of chemotherapy included younger age, histology (SCLC > NSCLC), increasing CS, increasing KPS and lack of co-morbidities. Of all chemo pts, 72% received it concurrently with RT. Surgery was performed in 23% of NSCLC pts. Only 3% of all pts were treated on cooperative group trials or IRB-approved institutional trials.
Conclusions:This study establishes the general patterns of care for lung ca in RT facilities within the U.S. As supported by clinical trials, pts with limited stage SCLC and CS III NSCLC received chemo + RT > RT alone. Further improvements in staging, smoking cessation and increased accrual to clinical trials must be encouraged. (supported by NCI Grant CA 65435).
Supported by NCI grant #CA65435.
Label #0162: Lee, W., Moughan, J., Owen, et. al.
Purpose/Objective:To provide descriptive information on a representative national sample of prostate cancer patients treated with prostate brachytherapy (PB) in calendar year 1999.
Materials/Methods:A random survey for the Patterns of Care Study in Radiation Oncology of 58 facilities treating prostate cancer patients in 1999 in the United States was performed. A weighted sample of 36,496 prostate cancer patients was included in the 1999 survey. The main outcome measures were the clinical characteristics of men prior to treatment and the technical characteristics of PB. Patients were classified into three prognostic groups according to T stage, pretreatment PSA and Gleason score.
Results:A weighted sample of 13,293 patients (36%) was treated with PB. Compared to patients treated with external beam radiation therapy alone (EB), patients that received PB were significantly younger (mean age: PB 67.7, EB 70.8; p=0.0006). African-American and Hispanic men (21%) were less likely to be treated with PB compared to white men (41%) (p= 0.0958). The vast majority of men treated with PB received low dose rate permanent seeds (89%). Fifty-four percent of men received PB monotherapy (PBM) and the remaining forty-six percent were treated with EB in addition to PB (EBPB). The mean PSA for the PB group was 9.9 (PBM group 7.1, EBPB group 13.2; p = 0.0008). Of the patients treated with PB, 54% were categorized as low risk, 33% were categorized as intermediate risk, and 13% were categorized as high risk. Patients with more advanced disease according to the risk categories were more likely to receive EBPB. The addition of EB to PB was used in 20%, 68% and 95% of patients in the low, intermediate and high risk categories respectively (p = 0.004). Androgen deprivation therapy (ADT) was used in 40% of patients treated with PB and was more likely to be used in patients with more advanced disease. Of the men treated with low dose rate PB, 59% were treated with I-125 and 41% were treated with Pd-103. Iodine-125 was more likely to be used in men treated with PBM (71%) and Pd-103 was more likely to be used in men treated with EBPB (58%) (p=0.22). The median prescription doses according to isotope are as follows (I-125: PBM 144 Gy, EBPB 108 Gy; Pd-103: PBM 115 Gy, EBPB 90 Gy). Postimplant dosimetry was documented in 61% of cases treated with low dose rate PB, 46% of cases used CT imaging.
Conclusions:PB is used in approximately 36% of men treated with radiation therapy in the United States. The mean age of the men treated with PB is younger than the population of men treated with EB alone. Nearly one-half of men receiving PB also receive EB. EB is used more frequently in men with higher risk disease, but even in men with low risk disease, 20% receive EB in addition to PB. ADT is used in 40% of PB patients. Techniques and prescription doses are consistent with published guidelines.
Supported by NCI grant #CA65435.
Label #0163: Eifel, P., Moughan, J., Erickson, B., et. al.
Purpose/Objective:To determine the influence of research findings and evolving technology on the practice of radiation therapy in patients with carcinoma of the cervix.
Materials/Methods:Radiation facilities were randomly selected from two strata: those that treated <500 and those that treated >=500 cancer patients per year. Each facility submitted a list of eligible patients treated between 1996 and 1999. Patient records were randomly selected from eligible patient records at each institution. National estimates were made using weights that reflected the relative contribution of each institution and of each patient within sampled institutions. A total of 442 patients' records were reviewed in 59 facilities to obtain data about their characteristics, evaluation, tumor extent, and treatment. Data were compared with those obtained from the 1992-1994 survey.
Results:The median age of patients treated for intact cervical cancer was 56 years; 58% were white, 10% Hispanic, and 28% black. 83% had squamous cell carcinomas. Overall, 40%, 28%, and 32% of patients treated for intact cervical cancer had Stages IA-IIA, IIB, or III-IVA disease, respectively. CT continued to be the most common method of nodal evaluation, obtained in 78% of patients; 14% had lymphangiogram, 12% had MRI and 12% had surgical evaluation. 93% of the patients treated with curative intent were treated with a combination of external beam and brachytherapy. Of those who had brachytherapy, 76% had LDR intracavitary (ICRT), 12.4% had HDR ICRT, 6.1% had LDR interstitial, 2.2% had HDR interstitial, 1.2% had combinations of HDR and LDR, and 2% had incomplete information. The percentage of patients treated with HDR vs. LDR had not increased significantly since the 1992-1994 survey (P = 0.6). 70% of patients treated with LDR brachytherapy had 2
or more applications. The median dose to Pt A was 81.5 Gy for patients who had external beam +LDR brachytherapy. Patients who had HDR brachytherapy had a wide variety of fractionation schemes: 18% had <=3, 15% had 4, 36% had 5, 20% had 6-8, and 10% had 9-12 fractions. The median duration of treatment was 57 days for patients who had LDR and 58 days for patients who had HDR brachytherapy. Four facilities had no eligible patients; 21 of the remaining 55 treated 8 or fewer eligible patients during the 4-year study period. From the sample data, it is estimated 16.4% of patients were treated in facilities that treated <=2 patients with intact cervical cancer per year; in the 1992-94 survey 31% of patients were treated in facilities treating <= 2 patients/year (p =0.09). Of patients who had external beam and brachytherapy, 17% had these treatments in two different facilities; most had external beam in a smaller facility and brachytherapy in a larger one. Thirty-three percent of patients treated in 1996-2000 received chemotherapy; of these 98% received concurrent chemotherapy. Most patients of the patients treated with chemotherapy received cisplatin alone (59%) or cisplatin with one or more other drugs (11%); 16% received concurrent hydroxyurea. The proportion of patients receiving chemotherapy was not significantly different from that observed in the 1992-94 survey. However, in the last year of the survey (1999), 63% of patients received chemotherapy compared with 19%, 26%, and 25% in the first three years of the survey, respectively. Details of chemotherapy administration were rarely documented in the radiation oncologist's or in surveyed hospital records; 67% of records failed to document the number of
chemotherapy cycles that were administrated with radiation.
Conclusions:The sharp increase in use of concurrent cisplatin-based chemotherapy in 1999 suggests rapid application of results of randomized trials. However confirmation of chemotherapy administration is not well documented in radiation oncologists' records. There continues to be considerable heterogeneity in practice patterns, particularly in the use of brachytherapy. However, the proportion of patients treated with HDR brachytherapy has not increased significantly since the last PCS survey.
Supported by NCI grant #CA65435.
Label #0170: Langer, C., Moughan, J., Movsas, B., et. al.
Purpose/Objective:In LD-SCLC, combined modality therapy has emerged as the standard of practice in good performance status (PS) patients (pts). Pignons meta-analysis NEJM 1992; 327:1618-24) showed that combination chemotherapy and thoracic radiation (XRT) in LD-SCLC yielded an absolute 5.4% increase in 3-yr survival vs. chemotherapy alone. Concurrent chemoradiation upfront has generated the highest survival rates. (Murray JCO 1993; 11: 336-344; Jeremic JCO 1997; 15: 893-900; Takada JCO 2002; 20: 3054-60) In stage III NSCLC, six separate studies have shown therapeutic superiority for combination chemotherapy and XRT vs. RT alone (Dillman NEJM 1990; 323:940-945; Sause Chest 2000; 117:358-364; Schaake-Koning NEJM 1992; 326:524-530;Jeremic Cancer 1993; 71: 3732-3736), and recent literature suggests a therapeutic advantage for concurrent chemoradiation vs. chemotherapy (rt arrow) XRT (Curran ASCO 2000; 19: 484a; Furuse JCO 1999; 17: 2692-2699; Zatloukal ASCO 2002;A-1159). Data are less secure regarding the role of chemotherapy in stage I and II NSCLC.
Materials/Methods:A stratified two-step cluster sampling technique was used for data collection. 541 individuals diagnosed between 1998 and 1999 with lung cancer, either LD-SCLC or stages I, II, and III NSCLC were sampled from 58 institutions, which gives a weighted sample size (wss) of 42,335 patients. All pts received XRT and had KPS >=60. We determined the percentage who received chemotherapy; the nature of hemotherapy and its timing with respect to XRT. SUDAAN statistical software was used to allow the incorporation of the design elements and weights that reflect the relative contribution of each institution and each patient in the analysis.
Results:Of 72 (wss=6,138) pts with LD-SCLC, 100% received XRT and 95% received chemotherapy (CT); 66% received concurrent (con) CT and XRT, of whom 52% also received CT pre XRT, and 45% received CT post XRT as well; 63% received sequential CT (rt. arrow) XRT+ con; and 38% received some CT after XRT. 52% received cisplatin (DDP), and 38% received carboplatin (CBDCA); 73% received VP-16, while 10% received paclitaxel. Of 469 pts (wss_36,197) with NSCLC, 52% received CT, including 30% with stage I disease, 48% with stage II NSCLC, 60% with stage III NSCLC, and 50% with unknown stage. 39% received sequential CT XRT+CT, of whom 49% received CT pre XRT only. 74% received con CT & XRT; and 27% received posterior CT, of whom 84% also received con CT/XRT. 4% received some CT in the pre-op setting and 9% in the post-op setting. 12% received DDP-based therapy, while only 13% and 7% received VP-16 or vincas respectively; 67% received CBDCA. 72% received taxanes, of whom 96% received paclitaxel. Gemcitabine was given to 3% of NSCLC patients.
Conclusions:Combined modality therapy is typically employed in the therapy of LD-SCLC and LA-NSCLC. The majority of those treated for SCLC receive concurrent CT/XRT, while nearly 3/4 of those treated with CT and XRT for LA-NSCLC received concurrent CT/XRT. Current practice generally matches evidence-based literature, although a significant % of practitioners substitute CBDCA for DDP in both venues and use paclitaxel in lieu of vincas or etoposide in NSCLC.
Supported by NCI grant #CA65435.
Label #0174: Kachnic, L., Moughan, J., Thomas Jr., C., et. al.
Purpose/Objective:A Patterns of Care Study (PCS) was conducted to evaluate the standards of practice for patients (pts) receiving radiation therapy (RT) for rectal cancer from 1992–1994 in the U.S. Specific aims of this study were to examine the evaluation, treatment schemes and outcome during this time period, and to compare these results to the original 1988–89 PCS dataset.
Materials/Methods:A national survey was conducted using a two-stage cluster sampling technique. Information was collected on 515 pts (weighted sample size = 21,675) with rectal cancer who received adjuvant pelvic RT at 57 facilities (36% cases from academic centers). Follow-up data was collected at site visits and subsequently by mail survey. Pts were staged according to 1988 AJCC criteria; Gunderson-Sosin, MAC system was used for tumor sub-staging. Eligibility criteria: pelvic RT between 1992–1994, no evidence of distant metastasis, adenocarcinoma histology, no prior/concurrent malignancies, no prior pelvic RT, and no residual disease after surgery. Statistical analysis was performed using SUDAAN software to accurately reflect PCS sampling technique. For comparison, the original 1988–89 PCS rectal survey results of 406 pts were subjected to the same statistical procedures.
Results: Median (med) age was 65 yrs. Staging work-up included CT (16%), EUS (4%) and MRI (0.3%). EUS was not considered in the original survey. Surgery consisted of 15% LAR, 27% LAR with coloanal anastomosis, 34% APR and 11% local excision (LE). There was no difference in the use of sphincter-preserving surgery (SPS) [LAR + coloanal or LE] compared to 1988 – 89 (52.9% vs. 50.7%, p = 0.66), however SPS approaches increased at academic centers (AC)[53% vs. 42%, p = 0.20]. Adjuvant therapy consisted of preop RT 7%, preop chemoRT (CRT) 8%, postop RT 13% and postop CRT 56%. There was increased use of preop CRT compared to ’88-89 (8% vs. 3.5%, p = 0.12) with a marked shift towards preop CRT at AC (27% vs. 4%, p = 0.01). The chemotherapy agents most commonly used were 5-FU (76%) and 5-FU + levamisole (7%). Use of infusional 5-FU was surprisingly less in the recent dataset (17.5% vs. 26.5%, p = 0.15), although increased at AC (36.9% vs. 22.4%, p = 0.35). RT planning and delivery incorporated an increase in: daily treatment with 3– 4 fields (91.8% vs. 84.1%, p =0.18); slightly higher med total dose (TD) (1992–94: pelvis=45Gy, boost=9Gy; 1988 – 89: pelvis=45 Gy, boost=7.8Gy; TD <50Gy in 31.4% vs. 44.3% in ’88 – 89; p = 0.0512); prone treatment position (79.8% vs. 56.9%, p=0.0018); and false table tops (11.1% vs. 3.8%, p = 0.14). Small bowel contrast was used similarly, 30%, in both surveys. Med f/u for this PCS was 2 yrs (range 9d–10yrs). For patients with 5+ yr f/u, med survival appeared to be TNM sub-stage dependent and without significant improvement in comparison to the prior survey: T3N0, T1-2N1 = 7.7 yrs, T4N0, T1-2N2, T3N1 =7.2 yrs, and T3N2, T4N1, T4N2 = 6.8 yrs. Of 453 pts from the combined surveys with 2+ yr f/u (med 5.1 yrs), 56% of pts are deceased from the ’88 – 89 PCS (with 27% recurrences) vs. 31%,’92–94 (with 17% recurrences) (p = 0.001). Multivariate analysis (MVA) of the 1988 – 89 PCS (med f/u = 4.9 yrs, range 12d–14 yrs) showed significantly poorer OS for stage III pts (p = 0.0355) and significantly improved DFS for pts receiving trimodality therapy (p = 0.0204). MVA failed to identify significant predictors for OS for the 1992–94 PCS; however, stage III pts had significantly poorer DFS (p = 0.0492).
Conclusions:Comparisons from these two PCS rectal cancer datasets suggest an increase in: 1) preop RT-based strategies overall and significantly within AC; 2) delivery of > 50 Gy RT, likely secondary to improved bowel exclusion techniques; 3) sphinctersparing surgery within AC, although unchanged overall; and 4) CRT with infusional 5FU among AC. Median survival appeared to be unchanged in this analysis (continued rationale for the incorporation of novel chemotherapeutic agents for CRT), and survival by TNM sub-stage failed to validate the risk groups from the larger Intergroup pooled analysis (Gunderson. IJROBP 2002).
Supported by NCI grant #CA65435.
Label #0175: Suntharalingam, M., Moughan, J., Coia, L., et. al.
Purpose/Objective:The national practice for patients receiving radiation therapy for carcinoma of the esophagus was evaluated by a Patterns of Care Study conducted for patients treated from 1996-1999. This report describes the outcomes of these patients and compares these results to updated PCS data obtained during the 1992-1994 survey, in order to document whether previously reported changes in practice have resulted in improvements in survival for this patient population.
Materials/Methods:A national survey was conducted at 59 institutions in a stratified random sample selected from a master list of radiation therapy facilities throughout the United States. A stratified two-step cluster sampling technique was used for data collection. 414 patients were sampled from the 59 institutions, which gives a weighted sample size of 11,340 patients nationwide. Eligibility criteria included either definitive or adjuvant RT delivered between 1996-1999, no evidence of distant metastases, squamous cell or adenocarcinoma histology, KPS > 60, tumors in the thoracic esophagus, and no prior malignancies within the previous 5 years. Patient, tumor, and treatment characteristics were evaluated SUDAAN statistical software was used to allow the incorporation of the design elements and weights that reflect the relative contribution of each institution and each patient in the analysis. Multivariate comparisons of survival times were made using the Cox proportional hazards model.
Results:The median age of patients was 64 years, 77% were male, and 23% female. KPS was > 80% in 85% of pts. Adenocarcinoma was diagnosed in 51% of pts and squamous cell carcinoma in 49%. 16% were clinical stage I (AJCC 83 system), 39% CS II, and 33% CS III (12% unknown). Concurrent chemoradiation continued to be the most commonly employed treatment strategy (97%). The median follow up for the entire group was 7 months. 91 patients had died (55% of these from esophageal cancer) at time of analysis. Significant variables in the multivariate analysis of survival times included clinical stage and treatment approach. Patients with clinical stage III disease had a higher hazard risk of death as compared to other stages of disease (clinical stage III risk ratio [RR] 2.66 vs. clinical stage I, p =0.004), whereas those treated with concurrent chemoradiation followed by surgery had a decreased risk of death compared to other treatment schemes (chemoradiation plus surgery RR 0.33 vs. chemo/RT only, p <0.0001). Histology and size of treatment center were not significant. An updated
analysis of the original survey performed between 1992-1994 confirmed these findings (median follow-up = 10 months). Multivariate analysis revealed that clinical stage I (p< 0.0001), treatment approach with trimodality therapy (p = 0.0013), and type of treatment center all predicted for improved overall survival times. When data from both surveys were combined these predictive variables continued to be significant. Patients presenting with clinical stage III disease had a RR of 2.26 as compared to other stage I patients (p=0.0004). Those patients receiving chemo/RT followed by surgery had an RR of 0.58 as compared to those receiving chemo/RT only (p=0.004).
Conclusions:The outcome results of the 1996-99 PCS survey confirm the findings of the 1992-94 study. Concurrent chemoradiation continued to be the most commonly utilized treatment approach during the time period studied. The observation that patients undergoing surgical resection following concurrent chemoradiation have a decreased hazard or chance of death (i.e., improved overall survival) compared to other treatment schemes highlights the need for a randomized trial comparing these strategies.
Supported by NCI grant #CA65435.
Label #0173: Movsas, B., Moughan, J., Owen, J., et. al.
Purpose/Objective:To identify factors significantly influencing accrual to clinical protocols by analyzing Patterns of Care Study (PCS) surveys of 3047 randomly selected radiation (RT) patients.
Materials/Methods:PCS surveys from disease sites studied between 1992-94 and 1996-99 (breast cancer [1080 pts], prostate cancer [1149 pts], esophageal cancer [818 pts]) were analyzed. PCS is an NCI-funded national survey of randomly selected RT institutions in the U.S. Patients with non-metastatic disease who received RT as definitive or adjuvant therapy were randomly selected. To determine national estimates, sample size was weighted by the relative contribution of each institution and patients within each institution. Accordingly, 3,047 patient records were reviewed by trained research associates; the weighted sample size was 132,890. Data regarding participation in clinical trials were recorded. The following factors (age, gender, race, type
of insurance and practice type of treating institution [academic or not]) were studied by univariate and multivariate analyses.
Results:Overall, only 2.7% of all patients were accrued to clinical protocols, consistent with national estimates. Of these, 56% were enrolled on IRB-approved institutional trials and 44% on NCI collaborative group studies. On multivariate analysis, practice type (p=0.009) and race (p=0.012) were highly significant predictors of patient accrual. On univariate analysis, 9.4% of patients treated with RT at an academic practice were accrued to trials vs. 1.7% at non-academic sites (p=0.011). Yet, the vast majority (87%) of all patients were seen at non-academic sites. White patients enrolled onto studies more than African Americans (2.8% vs. 0.8%, p=0.024). A higher percentage of patients with esophageal cancer (9.8%) were accrued compared to either breast cancer (2.2%) or prostate cancer (1.98%), p=0.03, respectively. Younger age demonstrated a trend towards increased enrollment (3.5% for <70 yrs vs. 1.5% for >=70 yrs, p=0.058). However, gender (male 3.4% vs. female 2.3%, p=0.4) and type of insurance were not predictive (Medicare 1.9%, HMO 1.8%, Private [eg. BCBS] 3.2%, other 5.2% [private vs. all others, p=0.79]). A trend towards increased accrual over time was observed (1.5% from '92-'94 vs. 4.2% from '96-'99, p=0.058).
Conclusions:Practice type and race significantly influence enrollment onto clinical oncology trials. This suggests that increased communication & education regarding protocols, particularly focusing on physicians in non-academic settings and minority patients, will be essential to enhance accrual. While there has been some increase in accrual during the 1990's, further improvement is critical.
Supported by NCI grant #CA65435.
Label #0180: White, J., Moughan, J., Pierce, LJ, et. al.
Background:There have been multiple consensus statements and / or practice guidelines published recently addressing the recommended use of post-mastectomy radiation (PMR) for breast cancer patients. The Patterns of Care Study (PCS) of the American College of Radiology (ACR) used data from its 1998-99 survey of PMR to determine what proportion of breast cancer patients treated prior to these guidelines met the subsequent criteria for treatment with PMR in order to establish whether distribution of the guidelines will significantly change radiation practices.
Methods:A PCS research associate collected data for breast cancer patients treated with PMR in 1998-99 from 55 institutions selected from a stratified random sample of the national practice. A weighted sample of 13,720 patient records was obtained from a sampling of 405 records. Indications for PMR were summarized from statements by the American Society of Therapeutic Radiology and Oncology, ACR, National Cancer Institute, and American Society of Clinical Oncology. PMR was considered according to guidelines (PMR-G) if one of the following were present: >=4 nodal metastases, stage III, or T-3 N0 disease. PMR was consistent with guidelines if it included treatment to the chest wall + supraclavicular (SCL) nodes without a separate axillary field.
Results:Breast cancer patients treated with PMR in 1998-99 had a mean tumor size of 3.5 cm and a mean of 4.5 axillary nodal metastases. An average of 16 nodes were recovered from the dissection specimen. Systemic therapy was given to 96.5%: chemotherapy 92% and Tamoxifen 55%. Thirty nine % had >=4 nodes positive, 6.7% were T-3 N-0, and 14.5% stage III. PMR-G indications were found for 52%. Of the 48% that did not meet the PMR-G criteria, 39% were node negative, 59% had 1-3 positive nodes. Patients in the PMR-G group had a larger mean tumor size (4.4 cm vs. 2.6 cm, p=0.0016) and greater mean number of nodes positive (7.6 vs. 1.0, p<0.0001). PMR-G was less frequent at institutions treating < 500 total patients/ year (p=0.0053) and was unaffected by menopausal status (p=0.2779). PMR included treatment to the chestwall for 100% and SCL for 76%. Of those receiving SCL radiation, 47% were treated with a separate axillary field.
Conclusion:Half the patients treated with PMR in 1998-99 had an indication for treatment according to multiple subsequently published treatment guidelines. This represents a baseline from which either penetration of the guidelines into radiation practice can be gauged or the need for revision of the guidelines can be judged to better reflect prevailing practice and/ or include emerging data.
Label #0183: Movsas, B., Moughan, J., Owen, J., et. al.
Background:The purpose of this study was to identify key factors influencing accrual to clinical oncology trials.
Methods:Patterns of Care Study (PCS) surveys from sites studied between 1992-94 and 1996-99 (prostate cancer [1149 patients], breast cancer [1080 patients], esophageal cancer [818 patients]) were analyzed. PCS is an NCI-funded national survey of randomly selected radiation (RT) institutions in the U.S. Patients with non-metastatic disease, who received RT as definitive or adjuvant therapy, were randomly selected. To determine national estimates, the sample size was weighted by the relative contribution of each institution and patients within each institution. Accordingly, 3,047 patient records were reviewed by trained research associates; the weighted sample size was 132,890. The following factors (age, gender, race, type of insurance and practice type of treating institution [academic or not]) were studied by univariate and multivariate (MVA) analyses.
Results:Only 2.7% of all patients were accrued to clinical oncology protocols, consistent with national estimates. On MVA, practice type (p=0.009) and race (p=0.012) were significant predictors of accrual. On univariate analysis, 9.4% of patients treated with RT at an academic practice were accrued vs. 1.7% at non-academic sites (p=0.011). Yet, the vast majority (87%) of all patients were seen at non-academic sites. White patients enrolled onto studies more than African Americans (2.8% vs. 0.8%, p=0.024). Younger age demonstrated a trend towards increased enrollment (3.5% for <70 yrs vs. 1.5% for >70 yrs, p=0.058). However, gender and type of insurance were not predictive. A trend towards increased accrual over time was observed (1.5% from '92-'94 vs. 4.2% from '96-'99, p=0.058).
Conclusion:Practice type and race significantly influence accrual onto clinical oncology trials. To enhance enrollment, increased communication & education regarding protocols is essential, particularly focused on physicians in non-academic settings and minority patients. Supported by NCI grant CA 65435.
Label #0190: White, J., Moughan, J., Pierce, et. al.
Purpose/Objective:Regional nodal irradiation (RNI) after mastectomy has been the subject of much debate and has generated multiple guidelines about which population of node-positive breast cancer patients benefit most and how radiation should be executed. It is unclear whether similar criteria for RNI are being applied following breast conservation (BC) and mastectomy. The purpose of this study was to utilize the Patterns of Care Study (PCS) to examine how RNI was implemented in Stage I-II cases for BC compared to postmastectomy (PM).
Materials/Methods:PCS used a two-staged stratified random sample to perform two separate surveys of breast cancer patients treated in 1998-99:353 survey cases (71,877 weighted sample size [wss]) who underwent radiation after BC surgery and 405 survey cases (wss = 13,720) who received radiation PM. The surveys were conducted simultaneously at 59 randomly selected institutions. For this study, the analyzed population is node-positive and any node-negative breast cancer case that received RNI from both surveys. This resulted in 53 survey (wss = 11,105) BC cases and 208 survey (wss = 7,020) PM cases. Multivariate analysis was done to compare what factors were associated with the type of RNI utilized.
Results:The populations that received RNI after BC and PM had statistically similar characteristics in terms of the distribution of ethnicity, principal payer, and menopausal status. Patients in the BC group who received RNI are slightly older: mean age was 59.3 yrs vs. 52.7 yrs in the PM group (p = 0.03). PM patients were more likely to be treated at higher volume facilities (treating >500 cases/year, p = 0.01), and academic facilities (p = 0.0003) than their BC counterparts. Similar size tumors were treated with a mean tumor size of 2.1 cm for BC vs. 2.6 cm in the PM group (p = 0.10). Axillary dissection was performed for 72% of the BC cases and 100% of the PM cases (p = 0.004) with similar numbers of nodes within the specimen, (17.9 vs. 17.3 respectively; p = 0.76). After dissection in the BC group 25.4% were node-negative, 56.1% had 1–3 positive nodes, and 18.5% had =4 positive nodes compared to 10.9% node negative, 40.3% 1–3 positive nodes, and 48.8% > = 4 positive nodes for PM (p = 0.06). The distribution of nodal metastases differed significantly between BC and PM for >=4 positive nodes, (p = 0.02). Extra capsular extension was more frequent in the PM cases, 30.1% vs. 11.1% (p = 0.056).
In the BC survey 47.4% of the patients with 1–3 and 90.6% of the >=4 positive nodes underwent RNI. When RNI was performed in the BC survey, it included a supraclavicular field (SCL) 80.2%, an axillary (AX) field 36.7%, and internal mammary field (IM) 4.9% as compared to the PM survey, when RNI included a SCL for 98.8% (p = 0.0495), AX 42.5% (p = 0.64), and IM 28.2% (p = 0.005). CT based treatment planning for RNI was used for 35.8% for the BC patients compared to 21.1% for the PM (p = 0.26). In the PM group, lower energy (<8 MV/MEV) was used more frequently for the SCL (p = 0.0495) and AX (p = 0.10) fields. There were no other significant differences between the BC and PM groups in the radiation type (electron or photons), dose, and where the dose was recorded. On multivariate analysis of the BC cases, a T-2 tumor was associated with receiving RT to the SCL field (OR 4.3, p = 0.04). In the PM survey, a T-2 tumor size was associated with RNI to the AX field (OR 2.9, p = 0.03). When the two surveys are combined for multivariate analysis, patients with positive nodes (p < 0.0001) and patients in the PM survey (p = 0.004) had a higher chance of receiving RNI to the SCL field than negative node and BC patients respectively. Conclusion:Patients with Stages I-II breast cancer who received RNI in 1998–99 after BC had similar size tumors but were less likely to have >=4 positive nodes compared to those treated PM. RNI after BC was significantly less likely to include the internal mammary nodes. Having had a mastectomy rather than BCT was associated with the use of RNI. Supported by NCI grant CA 65435.
Label #0189: Kachnic, L., Moughan, J., Thomas Jr., C., et. al.
Purpose/Objective:This retrospective analysis was performed on the PCS 1988–89s1992–94 RC datasets to determine if significant differences exist in the treatment approaches among patients (pts) managed at AC vs. NAC centers.
Materials/Methods:Two PCS national RC surveys were conducted using a 2-stage stratified sampling technique. Information was collected on 406 pts (weighted sample size=11,846)s515 pts (21,676) with RC who received adjuvant pelvic RT at 69 and 57 facilities within the ’88–89s’92–94 survey periods, respectively. 12% of pts were treated at AC in both surveys. AC were defined as an NCI-comprehensive cancer center (CCC) or had 2 or more radiation oncology residents in the ’88–89 PCS, and as a CCC or primary teaching hospital of a medical school in ’92–94. Pts were staged according to 1988 AJCC TNM criteria. Eligibility criteriasstatistical methodology per SUDAAN software, to accurately reflect PCS sampling technique, has been previously described. For comparisons within groups (i.e. AC ’88–89 vs. AC ’92–94) or between groups (i.e. AC vs. NAC), tests for associations were performed using Pearson chi-square statistics.
Results:In treatment comparisons between ACsNAC for all pts from both surveys, AC were significantly more likely to use preop chemoradiation (CRT) [22% vs. 7%, p < 0.0001], while NAC displayed a preference for postop CRT (52% vs. 41%, p = 0.0003). AC were significantly more likely to use chemotherapy (CT) [71% vs. 63%, p = 0.002], concurrent CRT with 5-FU (93% vs. 78%, p < 0.0001),sinfusional 5-FU (54% vs. 36%, p = 0.0001). There was a significant increase in the use of bowel-sparing RT techniques at AC (prone position, belly board, 3-field plans, > 6 MV energy, small bowel contrast; all p values < 0.01). Surgical procedure types were similar amongst both groups. Comparisons between survey periods (’88–89 vs. ’92–94) demonstrated a significant decline in the use of RT without CT, most notably at AC (preop RT 13% vs. 2%, p < 0.0001; postop RT 45% vs. 8%, p < 0.0001) in ’88–89 vs. ’92–94 respectively. The use of preop CRT significantly increased (4% ’88–89 vs. 32% ’92–94, p < 0.0001) in AC, with an increase in the use of postop CRT at NAC (41% ’88–89 vs. 58% ’92–94, p < 0.0001). A dramatic change in the role of CT occurred, with a marked increase in the use of any CT (39% ’88–89 vs. 89% ’92–94, p < 0.0001) in AC, and a significant increase in the use of concurrent CRT with 5-FU (54% vs. 96%, p < 0.0001) at NAC. Surgical procedures also changed, as local excision rates were significantly increased at AC over the survey periods (4% ’88–89 vs. 19% ’92–94, p = 0.0009). Additionally, a significant increase in the use of bowel-sparing RT techniques was demonstrated between survey periods for both AC and NAC (prone position, 3-field plans, daily treatment of all fields, belly board; all p values < 0.01). Interestingly, the proportion of pts receiving a total dose (TD) < 54 Gy increased in AC (71% vs. 81%, p = 0.005), while the TD > 54 Gy increased (20% vs. 31%, p < 0.0001) at NAC.
Conclusion:Treatment center comparisons from these two PCS RC datasets suggest: 1) only AC have significantly endorsed CT and preoperative CRT-based strategies over the survey periods. This increase in use of CRT strategies may have limited the ability for North American cooperative groups (NSABP R-03, RTOG-9401/INT-0147) to successfully define the optimal CRTssurgery sequence for T3/4 and/or N+ RC; 2) Sphincter-sparing local excision procedures were significantly increased at only AC; 3) AC > NAC have significantly increased the use of bowel-sparing RT planning and delivery techniques; however, 4) delivery of > 54 Gy RT was more common at NAC. Supported by NCI Grant CA 65435
Label #0188: Thomas Jr., C., Moughan, J., Kachnic, L., et. al.
Purpose/Objective:A retrospective exploratory analysis was performed on the PCS 1988–89s1992–94 RC datasets. The specific aims of the present analysis were to determine if differences exist in the outcome among patients (pts) managed at AC vs. NAC centers. Significant observations may be helpful in the stratificationsdescription design of future prospective, multi-institutional clinical trials for RC.
Materials/Methods:Two PCS RC national surveys were conducted using a 2-stage stratified sampling technique. Information was collected on 406 pts (weighted sample size = 11,846)s515 pts (weighted sample size = 21,676) with RC who received adjuvant pelvic RT at 69 and 57 facilities within the 1988–89s1992–94 survey periods, respectively. 12% of pts were treated at AC in both surveys. In the 1988–89 survey, an AC was defined as an NCI-comprehensive cancer center/main RTOG member or had >= 2 radiation oncology residents. The 1992–94 survey defined an AC as being either a NCI-comprehensive cancer center or primary teaching hospital of a medical school. F/u data was collected at site visitsssubsequently by mail survey. Pts were staged according to 1988 AJCC TNM criteria. Eligibility criteriasstatistical methodology per SUDAAN software to accurately reflect PCS sampling technique has been previously described. For comparison within groups (i.e. AC 1988–89 vs. AC 1992–94) or between groups (i.e. AC vs. NAC), the Cox proportional hazards model was used.
Results:In outcome comparisons between ACsNAC for all patients from both datasets, OS (HR 1.35, p=0.05)sDFS (HR 1.37, p < 0.03) were significantly improved in AC after adjusting for treatment approach and stage. Local-regional recurrences (LRR) were not significantly different (AC-HR1.20, p = 0.55). These observations were maintained in the cohort of pts with 2+ and 5+ yrs f/u. Outcome analysis of survey periods (1992–94 vs. 1988–89) did observe a modest, although non-significant, improvement in OSsDFS in the AC 1992–94 vs. AC 1988–89 PCS (OS HR 1.66, p = 0.11; DFS HR 1.15, p = 0.60), as well as in OS for NAC from the 1992–94 survey vs. NAC 1988–89) (HR 1.20, p = 0.11). Interestingly, the LRR risk was lower in the 1988–89 vs. 1992–94 survey for NAC pts (HR 0.63, p = 0.03) after adjusting for treatment approach and stage. These results were maintained in pts with 2+ and 5+ yrs f/u. Pts at AC receiving less than trimodality therapy had statistically significant worse OS (HR 2.36, p = 0.005) and DFS (HR 2.79, p = 0.0005) than pts who received trimodality treatment after adjusting for survey and stage. There was no statistical significance, however, for LRR. Pts at NAC receiving less than trimodality therapy did not experience similar changes in OS or DFS, but did in LRR (HR 2.05, p = 0.013).
Conclusion:This PCS analysis demonstrates that patients treated at AC had superior OSsDFS results with similar local recurrence patterns. These survival disparities between ACsNAC are in large part due to the increased use of trimodality therapy at AC during these study periods. The higher risk of LRR in the 1992–94 PCS at NAC may be a reflection of the increased RT dose in this cohort (> 54 Gy) at the time of the survey. While this analysis is unable to determine whether the aforementioned outcome disparities among AC vs. NAC still persist in the U.S., the importance of adjuvant chemoradiation and RT dose continue to be addressed in ongoing national trials. Supported by NCI Grant CA 65435
Label #0187: Coia, L., Moughan, J., Suntharalingam, M., et. al.
Purpose/Objective:Patients with esophageal cancer who do not have esophagectomy as initial treatment must undergo endoscopic ultrasound staging (EUS) to be staged by the current AJCC staging system. Since most patients do not undergo EUS prior to therapy, a modified 1983 AJCC staging system has been used by the PCS (Table 1). This study examines the merits of EUS (AJCC 1997) and PCS staging for patients with esophageal cancer treated initially with chemoradiation.
Materials/Methods:PCS conducted nationwide process and outcome surveys in 1992-1994 and 1996–1999 on 818 patients with esophageal cancer who were treated with radiation as a component of their definitive management. Follow-up information was available for 802 of these patients, which represents a weighted sample size (wss) of 10,180 patients nationwide. Of these patients, 82 (wss=955) underwent pre-chemoradiation EUS staging and form the basis for this retrospective investigation. Statistical analysis was performed using the SUDAAN statistical software to allow the incorporation of the design elements and weights that reflect the relative contribution of each institution and each patient in this study. Kaplan-Meier estimates of the survivor function were obtained within defined stage levels. Multivariate and univariate comparisons of survival times were made using the Cox proportional hazards model.
Results:The clinical stage distribution for 82 patients per both PCS and EUS stage was: PCS I/II 29%, PCS III 71%; EUS I/II 48%, EUS III 52%. Of PCS I/II, 14% were EUS III and of PCS III, 32% were EUS I/II. The hazard ratios (HR) for death increased by PCS stage for all 802 patients (wss=10,180), (I=1.0, II=1.16,III=1.93) (III vs. I, p < 0.01).Three-year survival for all patients was 44%, 36% and 18% for PCS stage I, II,sIII respectively. The HR for death demonstrated a trend for the 82 patients (wss = 955) by EUS stage (EUS I/II = 1.0, III = 2.3; p = 0.12), as well by PCS stage (PCSI/II=1.0,III=2.5;p = 0.25). For patients staged by both PCS and EUS, significant survival differences were seen for patients with PCS III, EUS I/II (HR 0.46; p = 0.05) vs. PCS III, EUS III (HR 1.0). Three prognostic groups were identified: Group 1: PCS I/II, EUS any; Group 2: PCS III, EUS I/II and Group 3: PCS III, EUS III. Three-year survival was 83%, 61% and 29% for Groups 1, 2 and 3 respectively. Multivariate analysis of the components of each staging system indicated only tumor length <=5 cm (HR 1.0) vs. >5 cm (HR 10.3; p = 0.03)to be statistically significant.
Conclusion:For patients with esophageal cancer who were managed initially with chemoradiation, both the PCS staging system and EUS staging system (1997 AJCC)are of prognostic significance. EUS staging provides additional prognostic information for patients with PCS stage III esophageal cancer, and therefore should be strongly encouraged in addition to PCS staging.
Label #0186: Erickson, B., Eifel, P., Moughan, J., et. al.
Purpose/Objective:To analyze the details of brachytherapy (BT) practice in patients treated for carcinoma of the cervix in the USA from 1996-1999.
Materials/Methods:Radiation facilities were randomly selected from two strata: those that treated <500 or >=500 cancer patients/yr. Patients were randomly selected from lists of eligible patients treated at each facility. A total of 442 patients records were reviewed in 59 facilities to obtain data about their characteristics, evaluation, tumor extent, and treatment. National estimates were made using weights that reflected the relative contribution of each institution and of each patient within the sampled institutions. From our survey we estimate that 16,375 patients were treated in the U.S. during this study period. Unless otherwise specified BT practice was based on the 408 patients who had their BT or all their treatment at the surveyed facility.
Results:91.5% of patients underwent BT at the initial treating institution; 8.5% were referred to a second site for BT. 42% of U.S. facilities referred at least some patients to a second facility for BT. Of U.S. facilities that treated <=2 eligible patients per year, 61% referred all of their patients to a second facility for BT or treated with external RT alone; none of the U.S facilities with larger experiences (>2 eligible pts/yr) referred all their patients to a second facility for BT treatment but 28% referred some patients to an outside facility for BT. Overall, 94% of patients who completed treatment with curative intent received BT. Of patients who had BT, 77.8%, 13.3%, and 0.9% respectively were treated with low dose-rate (LDR), high dose-rate (HDR), or a combination of HDR and LDR BT; 7.9% had interstitial BT (5.7% LDR and 1.9% HDR, 0.3% Mixed). In facilities that treated >2 pts per year, 15.5% and 9.4% of BT procedures included HDR or interstitial respectively; in facilities that treated fewer patients, 3.4% had HDR BT and only 1.2% had interstitial BT. Patients treated with LDR intracavitary radiotherapy (ICRT) had 1 (27.8%), 2 (70%) or 3 (2.2%) implants. For curative patients who completed RT with LDR ICRT without hysterectomy the median BT dose to Pt A was 4031 cGy and the median total dose to Pt A was 8288 cGy. Patients were treated with HDR ICRT using a variety of treatment schedules using 1–2 fractions (fxs) (8%), 3–4 fxs (17%), 5–6 fxs (39%), 7–9 fxs (34%), or had 12 fxs (3%); fractions sizes were <500 (29%), 500– <600 (25%), 600 (28%), >600 (8%), or unknown (9%). For patients treated with HDR, the median total dose to Pt A (corrected for fraction size using an a/ß = 10) was 8580 cGy (5617–11606). At institutions treating <500 pts/yr, the % of patients receiving a BT dose <40 Gy was statistically higher than at institutions treating >=500 pts/yr (p < 0.0001). For LDR ICRT, 7.4% had no bladder and rectal doses recorded for all their implants, whereas in HDR ICRT, 75% had no bladder and rectal doses recorded for all their implants. The median total duration of RT was identical for patients who had HDR or LDR ICRT (57 days). For LDR at institutions treating <500 pts/yr, the % of patients with treatment duration >56 days was significantly greater than at institutions >=500 pts/yr (p = 0.0002). Of the patients who had LDR ICRT implants, 65% were treated using tandem and shielded FSD colpostats; other patients had mini ovoids (11%), cylinders (4%), Henschke (4%), or mixed applicators (15%). In contrast, of patients treated with HDR ICRT, 69% had tandem and rings, 18% FSD ovoids, 8% mini ovoids, and 6% other or mixed applicators.
Conclusion:The median duration of treatment and median Pt A dose were very similar for patients treated with HDR or LDR. Patients with HDR were treated using a variety of treatment schedules. Different applicator types were favored for LDR vs. HDR. 75% of patients treated with HDR had no BT bladder or rectal doses recorded, suggesting that full dosimetric calculations were only performed for the first fraction in many institutions. Facility size significantly impacted on referral to another institution for BT, BT dose, and treatment duration.
Label #0181: Johnstone, D., Moughan, J., Goldberg, M., et. al.
Purpose/Objective:A Patterns of Care Study was conducted to determine the national patterns of radiation therapy practice in patients with nonmetastatic lung cancer in 1998–99. An analysis of the cohort of patients who underwent surgery as part of their therapy was performed.
Materials/Methods:Patients with nonmetastatic lung cancer (KPS >= 60) who received radiation therapy as definitive or adjuvant therapy were randomly selected from 58 institutions. 541 patients were included in this study, which represents a weighted sample size (wss) of 42,335 patients nationwide. Of these pts, 131 (19.8%; wss = 8,398) underwent surgery as part of their therapy. Tests for associations between the surgical and nonsurgical patient groups were performed using Pearson chi-square statistics. Details of surgical resection, nodal staging, completeness of resection, and timing of radiation in relation to surgery were analyzed.
Results:PCS results have been crucial parts of the body of evidence leading to changes in practice. Helping stimulate clinical trials, measuring the penetration of clinical trials, assessing compliance with guidelines, showing when guidelines should be modified, and investigating disparities have been among the aims. As an example of helping to stimulate clinical trials, PCS showed that radiation dose affected tumor control and morbidity outcomes for prostate cancer patients. Those results led to dose escalation clinical trials in which PCS data was cited as preliminary data justifying trials of higher doses and was the main source of impetus to this research direction. Subsequent surveys showed shifts in the national practice to higher doses and use of conformal techniques. In breast cancer another type of PCS benefit is measurement of compliance with treatment standards and impact on guidelines. PCS results were cited as evidence justifying changes in Breast Cancer Clinical Practice Guidelines of the National Comprehensive Cancer Network (NCCN) and also stimulated changes in the Breast Imaging Reporting and Data System (BIRADS), standardized nomenclature for reporting breast imaging results, to assure that information necessary to guide treatment decisions was recorded.
Conclusion:Mounting societal demands for improvement in the quality of cancer care, ever increasing complexity of radiation therapy, and escalating use of multi-modality treatment make continuing to measure, report, and improve quality of care in radiation oncology more important than ever before. Methods must keep pace with new technologies and techniques in radiation therapy. Conducting this type of research remains crucial to patients and the profession.
Label #0204: Owen, J. and Wilson, J. Frank
Purpose/Objective:To describe the history and impact of a radiation oncology research initiative, unique in medicine, that has defined, measured, and reported quality of care for more than 30 years.
Materials/Methods:Based on the Donabedian model of quality assessment, the Patterns of Care Study (PCS) has analyzed three components crucial to quality care: structure, process, and outcome, by conducting Facilities, Process, and Outcomes Surveys. Starting with a complete census of radiation oncology facilities, the complex survey design allows calculation of national averages for patients treated with radiation oncology in selected disease sites. Over the years methods of data collection have evolved to incorporate modern information systems and future surveys will include image collection to allow assessment of modern technologies. Incorporating quality efforts of many other organizations has also allowed the project methodology to evolve and has led to a new focus – Quality Research in Radiation Oncology (Q-RRO). This will include definition and measurement of evidence-based quality indicators for patient care.
Results:Of the patients who underwent surgery, 126 (97.8%; wss = 8,214) had non-small cell lung cancer (NSCLC), compared with 85.5% in the overall study population. NSCLC pts are the focus of this analysis. Surgical patients were younger (mean age 60.6 vs. 68.9 yrs (p < 0.00001), had less weight loss (p = 0.11), had higher Karnofsky performance status (80+: 93.4% vs. 79%; p = 0.001), higher FEV1 (>2 liters: 20.3% vs. 3.3%; p = 0.03), and no difference in comorbidities (compared to nonsurgical pts). Only 51% of surgical patients had preoperative pathological staging of mediastinal nodes (any TBNA 21%, any mediastinoscopy or mediastinotomy 73%, VATS 1%), and no patients had sentinel node mapping. Surgery consisted of lobectomy or bilobectomy in 62.1%, pneumonectomy in 22.3%, wedge resection in 3.4%, no resection in 11.1%, and unknown in 1.1%. The mean number of lymph nodes resected was 7.4 per patient (range = 0 to 73). There was no significant difference in T stage between surgical and nonsurgical groups. Within the surgical group, clinical T staging (UICC 1997) was 41% T2, 26% T3, and 8% T4. Surgical patients had less N2/N3 disease than nonsurgical patients (29% vs. 47%; p = 0.002), and more N1 disease (28% vs. 7%; p = 0.0004). Less advanced nodal staging in the surgical group likely reflects patient selection for surgery as well as bias from pathological versus clinical staging data. Surgical TNM (UICC 1997) staging was 9.5% Stage I, 36% Stage II, 45.1% Stage III (9.5% unknown). At least 79% of patients received radiation in the adjuvant setting and 11% in the neoadjuvant setting (10% had unknown surgery sequence with RT). 50% of surgical pts had chemotherapy and of these, 12.3% had it preoperatively, 59.2% had it postoperatively, and 11.2% had it pre and postoperatively (17.3% had chemo with RT or unknown). In a logistic regression analysis, the following factors were found to be associated with patients having a higher chance of undergoing surgery: age <70 (OR 5.5, p = 0.0002 vs. age 70+), FEV1.0 >2 liters (OR 6.4, p = 0.02 vs. FEV1.0 > 2 liters), and N1 (OR 6.7, p = 0.008 vs. NX).
Conclusion:The national patterns of radiation oncology care concerning the use of local excision and postoperative radiotherapy demonstrate a lack of evidence-based clinical practice. The majority of patients receiving this approach had T2 and T3 tumors, moderate-poorly differentiated histology, and less than a full-thickness local resection. Such high-risk patients should be considered for definitive resection with/without adjuvant radiation-based treatment until prospective, randomized multi-center trials comparing radical surgery vs. local excision and adjuvant radiation-based therapy are performed.
Label #0203: Kachnic, L., Moughan, J., Minsky, B. et. al.
Purpose/Objective:Local excision (LE) of early stage rectal cancer preserves anal sphincter and normal sexual function. However, local recurrence after LE, despite the administration of adjuvant radiation-based therapy, remains a significant concern. Current literature suggests that local excision utilizing a full-thickness approach should be restricted to patients with stage T1 (and selected early T2) rectal cancers without high-risk pathologic factors. Two Patterns of Care Study (PCS) analyses were conducted to evaluate the national patterns of radiation oncology practice for patients receiving pelvic radiotherapy (RT) as part of adjuvant management for localized rectal cancer in the United States (1988-1989 and 1992-1994). The purpose of this analysis was to review national local excision practice patterns and the rates of failure in patients who had received LE and adjuvant RT for localized rectal cancer as per these two national PCS surveys.
Materials/Methods:Statistical analysis was performed using SUDAAN statistical software to accurately reflect the PCS two-stage sampling technique. SUDAAN analysis incorporates the design elements (the two stages of stratification) and weights that reflect the relative contribution of each institution and each patient. All estimates are based on the weighted sample size (wss). An unweighted sample size (uwss) of 406 patients (wss 11,846) with rectal cancer was identified for the 1988-89 PCS survey, and 515 patients (wss 21,675) for the 1992-94 PCS, from 69 and 57 facilities, respectively. Both academic centers and non-academic centers participated.
Results:A wss of 1,470 patients underwent LE procedures (uwss 80) from the two PCS surveys: full thickness in 43%; endoscopic resection, 13%; fragmented, 23%; fulguration, 1%; other, 10%; and unknown, 9%. Tumor stage was as follows: T1, 9% (MAC stage A); T2, 44% (B1 and C1); T3, 12%(B2 and C2); and 35% indeterminate. Tumor size was not available. The median distance of the tumor from the anal verge was 5.5 cm (range, 0.6 to 22 cm). Margin status was negative in 38%, positive in 23%, and indeterminate in 39%. Grade was well-differentiated in 16%, moderately-differentiated in 67%, poorly differentiated in 5%, and 12% unknown. Post-operative RT was administered in 90% of patients undergoing LE, with 10% receiving pre-operative RT (median dose 54 Gy, range 2 to 85). 5-FU based chemotherapy was administered in 31%, 66% of these patients receiving 5-FU concurrently with RT. Seven local recurrences have been documented with a median follow-up of 22 months (range, 1.5 months to 12 years).
Conclusion:Patients with nonmetastatic lung cancer who have undergone surgery and radiation therapy have less advanced disease clinically and pathologically than those undergoing radiation alone. Most radiation in this setting is given postoperatively, likely due to hilar and mediastinal nodal disease. Greater attention needs to be given to improved nodal staging preoperatively, and consideration of neoadjuvant treatment strategies in patients with mediastinal adenopathy.
Label #0208: Owen, J. and Wilson, J. F.
Background:To assess the overall quality of cancer care, efforts seek consensus quality indicators that crosscut health services. Although pilot programs have collected and analyzed relevant clinical data, they have been unable to collect radiation oncology (RO) data detailed enough to assess the quality of services or to inform key decision makers. Quality Research in Radiation Oncology (QRRO) conducted retrospective surveys of national practice since 1973 with major positive impact on the quality of practice through recursive processes.
Methods:From Donabedian’s model of quality assessment, QRRO analyzes crucial quality components by conducting Facilities, Process, and Outcomes Surveys. Survey design allows calculation of national averages for patients treated with RO and comparisons by key factors. Evolving data collection methods allow assessment of modern technologies. Methods start with definition and measurement of evidence-based quality indicators but allow greater detail and specification than most other quality measurement efforts.
Results:QRRO showed that radiation dose affected outcomes for prostate cancer patients. Higher radiation doses were associated with improved local tumor control rates and treatment techniques affected toxicity rates. These results, presented widely in numerous venues, stimulated dose escalation clinical trials. Trials conducted in the USA all used QRRO results as critical data, providing the major impetus to test new directions in dose escalation and new methods to target delivery more precisely. National practice shifted to higher doses and use of conformal techniques. The dataset is unique in providing cross-sectional information on practice patterns with a wide variety of treatment approaches from many institutions and sufficient details of treatment delivery to allow examination of questions about quality and effects of techniques.
Conclusion:Mounting societal demands for improvement in the quality of care, ever increasing complexity of radiation therapy, and escalating use of multi-modality treatment make continuing to measure, report, and improve quality of care in RO crucial to patients and the profession. Methods must keep pace with new technologies and techniques in radiation therapy. [Supported by NCI grant CA 65435.]
Label #0209: Owen, J. and Wilson, J. F.
Purpose:To assess the existing evidence base for measuring proposed Quality Indicators (QI) across all types of practice by analyzing data in an existing database for proposed QI.
Materials and Methods:A set of Quality Indicators (QI) was developed based on the National Comprehensive Cancer Network guidelines for breast cancer management. Published results of the ACR Patterns of Care Study (PCS) national survey database of patients receiving breast conservation therapy in 1998-99 was analyzed to assess compliance with these QI. This survey design used a two-stage stratified sample of radiation oncology practices and allows calculation of national averages of defined measures for patients treated in radiation oncology facilities throughout the country. Data was extracted from on-site chart reviews of medical and radiation therapy records.
Results:QI compliance results for the national practice include: for workup - mammography performed before excisional biopsy (97%), pathology review (96%), estrogen receptor status determined (83%), progesterone receptor status determined (81%), Her-2/neu test done and results known (31%) and for axillary staging - axillary dissection performed (82%). QI for locoregional treatment for patients with negative axillary nodes - radiation therapy (RT) to whole breast with boost (87%); for patients with 1-3 positive nodes - RT to whole breast + boost (51%), RT also to optional fields (43%); for patients with >= 4 positive nodes - RT to whole breast with boost to tumor bed and supraclavicular area (15%), RT also to optional fields (75%).
Significance / Conclusions: Benchmarking utilization patterns provides a foundation for assessing appropriateness of care in the future. Although not collected originally for the purpose of measuring Q I, the PCS database is a potentially rich data source that can be used to validate these measures. In addition to providing evidence on the usefulness of specific proposed Q I and identifying areas of practice needing improvement, this project allows recommendations for additional evidence-based QI. Supported by: Pennsylvania Department of Health, Tobacco Settlement Act 77-2001, Commonwealth Universal Research Enhancement (C.U.R.E.) for State Fiscal Year 2005.
This presentation uses published data from: Pierce, L., Moughan, J., White, J., Winchester, D., Owen, J. and Wilson, J. F.: 1998-1999 Patterns of Care Study Process Survey of National Practice Patterns Using Breast-Conserving Surgery and Radiotherapy in the Management of Stage I-II Breast Cancer. Int J Radiat Onco Biol Phys, 62(1):183-192, 2005.
Label #0210: Owen, J. and Wilson, J. F.
Research Objective:To assess the existing evidence base for measuring proposed Quality Indicators (QI) for breast cancer treatment by analyzing data in an existing database. To determine additional evidence needed to measure QI.
Study Design:QI were developed based on National Comprehensive Cancer Network (NCCN) guidelines for breast cancer management and were compared to published results of the Patterns of Care Study (PCS) database of patients receiving breast conservation therapy (BCT) in 1998-99. The survey data came from a national survey with a two-stage stratified sample of radiation oncology practices that allowed calculation of national averages of defined measures for patients treated in all types of radiation oncology facilities. Data was extracted from on-site chart reviews of medical and radiation therapy records.
Population Studied:Patients diagnosed with Stage I or II breast cancer, treated with BCT, who started radiation therapy (RT) in 1998 or 1999. Ineligibility criteria included: purely ductal carcinoma in situ or lobular carcinoma in situ, gross multicentric disease, bilateral breast cancer, mastectomy as primary treatment, previous breast RT, brachytherapy only, and prior or concurrent malignancies (exept nonmelanoma skin cancer).
Principal Findings:Compliance results for QI derived from NCCN guidelines for workup are: mammography performed before excisional biopsy (97%), pathology review (96%), estrogen receptor status determined (83%), progesterone receptor status determined (81%), Her-2/neu test done and results known (31%) and for axillary staging - axillary dissection performed (82%). QI for locoregional treatment for patients with negative axillary nodes are RT to whole breast with boost (87%); for patients with 1-3 positive nodes - the minimum recommended RT to whole breast + boost (51%), RT to optional fields in addition to the minimum (43%), thus patients receiving the minimum or more RT fields (94%); for patients with >= 4 positive nodes - the minimum recommended RT to whole breast with boost to tumor bed and supraclavicular area (15%), RT also to optional fields (75%), thus patients receiving the minimum or more RT fields (90%).
Conclusions: Benchmarking utilization patterns provides a foundation for assessing appropriateness of care in the future. Although not collected originally for the purpose of measuring QI, the PCS database is a potentially rich data source that can be used to validate these measures. In addition to providing evidence on the usefulness of specific proposed QI and identifying areas of practice needing improvement, this project allows recommendations for additional evidence-based QI.
Implications for Policy, Delivery or Practice: As policy makers implement pay for performance (P4P) programs, finding data sources to benchmark QI is important. The required level of detail exceeds most current administrative data, while clinical trials provide data only for a limited set of specific and usually experimental treatments. Well designed surveys can provide a data source with sufficient detail for a broad range of QI and define additional data requirements to allow use of administrative data to implement P4P programs.
Primary Funding Source: Pennsylvania Department of Health, Tobacco Settlement Act 77-2001, Commonwealth Universal Research Enhancement for Fiscal Year 2005. National Cancer Institute: CA 065435.
Label #0215: Piccoli, C., Owen, J. and Paulus, R.
Purpose: To evaluate variability in breast imaging services in Pennsylvania (PA).
Method and Materials: As part of the PCS/PA, two mailed questionnaires were sent to a sample of PA radiologists and to all breast imaging (BI) facilities in PA to determine facility ownership, affiliations, number and type of BI procedures, equipment, staffing and record-keeping practices. Responses from 70 clinically active radiologists and from 303 of 398 facilities were obtained.
Results: 12% of radiologists spent >=30% of their time and 4% spent ? 50% of their time in mammography which was a primary or secondary subspecialty of 10% and 3% of radiologists respectively. None of the radiologists responding had breast fellowship training. The weighted percentiles of all radiologists reading any mammography was 53%. By age group, weighted percentiles of radiologists reading mammography was 32% aged <45, 58% aged 45-54, 91% aged 55-64 and 32% aged >= 65. The mean number of mammograms per year read by all radiologists was 2757. Highest mean number of mammograms per year was in the >= 65 age group (5349). Of 297 facilities, 63% had 1 mammography unit, 23% had 2 units, 10% had 3 units and 4% had >= 4 units. 233 facilities were designated urban and 64 rural. Urban facilities were more likely than rural facilities to be freestanding (6% vs 0%), part of a general radiology office (22% vs 5%), affiliated with a medical school (17% vs 6%) or residency program (25% vs 16%). Rural facilities were more likely to be based in a hospital (80% vs 60%) or primary care physician office (13% vs 4%) and offer on-site stereotactic biopsy (47% vs 27%). Similar percentages of urban and rural facilities offered breast ultrasound, ultrasound guided biopsy and MRI.
Conclusion: In PA, one third of radiologists aged <45 and one third of radiologists >= 65 read mammograms, but the >= 65 age group read significantly more on average than the youngest group. Radiologists aged 45-64 read approximately 50% more mammograms than the other two groups combined. Rural mammography facilities were more likely to be hospital based or part of a primary care office and more likely to offer stereotactic biopsy than urban facilities.
Clinical Relevance/Application: By studying the range of approaches to providing screening mammography and diagnostic breast imaging procedures, inconsistencies in utilization and analysis of causes of variability can be analyzed.
Supported by the Pennsylvania Department of Health, Tobacco Settlement Act 77-201, Commonwealth Universal Research Enhancement (C.U.R.E.) program.
