Treatment outcomes of three-dimensional conformal radiotherapy for localized prostate carcinoma

A large community-based experience

Authors

  • Susan A. McCloskey M.D.,

    1. Division of Clinical Research, Valley Radiotherapy Associates Medical Group, El Segundo, California
    Search for more papers by this author
  • Nancy A. Ellerbroek M.D.,

    Corresponding author
    1. Division of Clinical Research, Valley Radiotherapy Associates Medical Group, El Segundo, California
    • Valley Radiotherapy Associates Medical Group, 880 Apollo St. #329, El Segundo, CA 90245
    Search for more papers by this author
    • Fax: (310) 335-4098

  • Lauren McCarthy,

    1. Division of Clinical Research, Valley Radiotherapy Associates Medical Group, El Segundo, California
    Search for more papers by this author
  • Arnold W. Malcolm M.D.,

    1. Division of Clinical Research, Valley Radiotherapy Associates Medical Group, El Segundo, California
    2. Department of Radiation Oncology, Valley Radiation Oncology Center, Tarzana, California
    Search for more papers by this author
  • May Lin Tao M.D., M.S.P.H.,

    1. Division of Clinical Research, Valley Radiotherapy Associates Medical Group, El Segundo, California
    2. Department of Radiation Oncology, St. John's Health Center, Santa Monica, California
    Search for more papers by this author
  • Robert C. Wollman M.D.,

    1. Division of Clinical Research, Valley Radiotherapy Associates Medical Group, El Segundo, California
    2. Department of Radiation Oncology, St. John's Health Center, Santa Monica, California
    Search for more papers by this author
  • Christopher M. Rose M.D.

    1. Division of Clinical Research, Valley Radiotherapy Associates Medical Group, El Segundo, California
    2. Department of Radiation Oncology, Providence St. Joseph Medical Center, Burbank, California
    Search for more papers by this author

Abstract

BACKGROUND

The current study documented the implementation of three-dimensional conformal radiotherapy and assessed the tumor control and toxicity of such treatment in a large, multisite community practice.

METHODS

The authors retrospectively reviewed their first 222 consecutive patients with clinically localized (N0) prostate carcinoma treated with a 6-field conformal technique from October 1993 through March 2000. Standardized target definitions, dose planning constraints, and gantry angles were utilized to develop the treatment plan. Patients were categorized by low, intermediate, and high risk. Low risk was defined as T1a–T2a disease, a Gleason score < 7, and prostate-specific antigen (PSA) level ≤ 10.0 ng/mL (n = 47 [21%]). Intermediate risk was defined as T2b disease, a Gleason score > 6, or PSA level > 10.01 ng/mL (n = 60 [27%]). High risk was defined as 2 of the above risk factors or as T3 disease, a Gleason score > 7, or a PSA level > 20 (n = 115 [52%]). Biochemical disease recurrence was defined in accordance with the American Society for Therapeutic Radiology and Oncology definition. Urinary and bowel toxicity were graded using the Radiation Therapy Oncology Group morbidity scoring system.

RESULTS

The median follow-up after radiotherapy for surviving patients was 47 months (range, 0–99 months). The 2 and 5-year actuarial biochemical control rates for all patients were 84% and 78%, respectively. Using logistic regression analysis, lower dose (< 75.6 gray [Gy] vs. 75.6 Gy; P = 0.006), higher risk group (P = 0.033), higher stage (P = 0.045), and higher PSA level (P = 0.001) were significantly associated with biochemical disease recurrence. Toxicity was not significantly correlated with a higher radiotherapy dose.

CONCLUSIONS

Dose escalation to 75.6 Gy using a 6-field conformal technique was feasible in the authors' community practice and resulted in acceptable toxicity and early biochemical outcomes. Cancer 2004. © 2004 American Cancer Society.

As prostate-specific antigen (PSA) screening has become widely utilized in the detection of prostate carcinoma, more tumors are being detected at an earlier, potentially curable stage. External beam radiotherapy is a well established treatment option for patients with localized prostate carcinoma. However, with the standard four-field or arc techniques, rectal and bladder toxicity has limited dose escalation to improve outcomes, as demonstrated in numerous retrospective and in one randomized trial.1–3 Computed tomography (CT)-based, three-dimensional radiotherapy planning is a commonly utilized technology for accurate tumor targeting and conformal treatment. Such improved precision permits dose escalation without significantly increasing toxicity, thus improving the therapeutic ratio.

Although three-dimensional conformal radiotherapy (3DCRT) is much utilized in both academic and community practice settings, thus far the literature primarily reflects outcomes from academic institutions.1–23 We, as a large, multisite community practice, instituted 3DCRT in 1993 and have a sample of patients who had their treatment planned with uniform techniques and dosimetric parameters by a single radiation oncologist. The primary aims of the current study were to demonstrate the implementation of 3DCRT with dose escalation for localized prostate carcinoma in a community-based practice and to report the toxicities and biochemical outcomes by risk groups.

MATERIALS AND METHODS

Between October 7, 1993 and March 30, 2000, 222 consecutive patients with clinically localized prostate carcinoma were treated with a 6-field conformal technique at 1 of 3 radiation facilities (Providence St. Joseph Medical Center, Providence Holy Cross Cancer Center, or Valley Radiation Oncology Center in Los Angles County, CA). Study data collection was approved by the institutional review board at all three treating facilities. Two hundred and fifty-two patients were identified after initial review of our registry of patients with prostate carcinoma treated with conformal radiotherapy. Patients were considered ineligible if they were not ultimately treated for diagnosed prostate carcinoma (n = 4), if they had prostate carcinoma that was not treated with 6 fields (n = 13), if they were diagnosed with cancer other than prostate carcinoma and were in the prostate registry in error (n = 6), if they were lymph node positive at the time of diagnosis (n = 1), if they were treated only for disease recurrence (n = 2), or if they had no treatment or follow-up data available (n = 6). The initial evaluation for all patients included evaluation of pretreatment PSA levels, histologic evaluation with Gleason's score, and American Joint Committee on Cancer clinical staging of disease.24 Further workup including bone scans, diagnostic CT and/or magnetic resonance imaging scans was done at the discretion of the treating physician.

Patients were categorized as low, intermediate, and high risk. Low-risk patients were defined as having T1a–T2a disease, a Gleason score < 7, and a PSA level ≤ 10.0 ng/mL. Intermediate-risk patients were defined as having T2b disease, a Gleason > 6, or a PSA level > 10.01 ng/mL. High-risk patients were defined as having any two of the intermediate risk factors or as having T3 disease, a Gleason score > 7, or a PSA level > 20 ng/mL.

The conformal treatment consisted of a six-field approach based on the previously described Memorial-Sloan Kettering technique.6 All patients underwent CT scan simulation with fabrication of a customized Aquaplast™ (WFR/Aquaplast Thermoplastics, Wynhoff, NJ) pelvic immobilization device. Clinical treatment volume 1 (CTV1) was defined as the prostate gland and seminal vesicles, and CTV2 was defined as the prostate only. The corresponding planning target volume (PTVs) were defined as the CTV plus 6 mm in the posterior dimension and 10 mm for all other dimensions (anterior, right, left, superior, and inferior). All volumes were contoured by one radiation oncologist (CMR), although patients were managed by one of several radiation oncologists in the practice. For all patients, 15-MV photons were used. The initial target was the prostate plus seminal vesicles in 119 patients and the prostate only in 103 patients. One hundred eight patients had a reduction off of the seminal vesicles, covering the prostate only, after a dose of 50.4 gray (Gy). Ninety-one patients, in general those determined to have a risk of seminal vesicle involvement of > 15% based on Partin tables,25 received 72 Gy to the prostate and seminal vesicles. One hundred forty-seven patients also received an additional boost to the prostate of 3.6 Gy using opposed lateral fields only, designed to exclude the rectum, for a total dose of 75.6 Gy. Seventy percent of the rectal volume received < 70 Gy and the maximum rectal dose was < 72 Gy.

Treatment was delivered in a uniform manner at all three facilities according to the medical group's policies. All fields were treated daily (Monday to Friday, except holidays) with 1.8–2.0-Gy fractions. The total prescribed dose to the prostate was 6840–7560 cGy. The dose was prescribed to the isodose line that encompassed 100% of the PTV. Weekly isocenter and biweekly cerrobend block verifications were performed with port films. Patients were seen weekly by the treating radiation oncologist during therapy. After therapy, patients returned every 3–6 months for the first 2 years and yearly thereafter. Acute toxicity was defined as toxicity occurring during radiotherapy or within the first 3 months from completion. Late toxicity was defined at toxicity occurring beyond 3 months after completion of treatment.

Biochemical disease recurrence was defined in accordance with the American Society for Therapeutic Radiology and Oncology (ASTRO) as three consecutive increases after reaching a posttreatment nadir.26 The date of biochemical disease recurrence was backdated to the midpoint between the posttreatment nadir and the first of three consecutive increases. Clinical disease recurrence was defined as clinical or radiographic evidence of local or distant disease. Genitourinary and rectal toxicity was graded using the Radiation Therapy Oncology Group (RTOG) morbidity scoring system.27 Patients were questioned before treatment regarding sexual potency. Only patients reporting potency were assessed in follow-up and given the designation of complete retention of potency, partial loss, or complete loss.

Actuarial biochemical control and actuarial overall and cause-specific survival were tabulated using the SAS proc lifetest.28 Logistic regression analysis was used to assess the statistical significance of risk group, stage, PSA level, dose, age, and hormone use with biochemical disease recurrence. A t test was used to assess the statistical significance of dose in the development of gastrointestinal (GI) and genitourinary (GU) toxicity. Chi-square statistical analysis was used to assess the relation between loss of sexual potency and radiotherapy dose. Patients lost to follow-up were censored at the time of their last follow-up visit.

RESULTS

Table 1 delineates the patients' demographic and clinicopathologic characteristics, treatment regimen, and length of follow-up data. The median patient age was 70 years (range, 43–85 years). One hundred eighty-two patients were white, 4 were African American, 26 were Hispanic, 3 were Asian, and 7 did not specify race. Clinical stage was T1 in 78 patients (1 T1a, 4 T1b, and 73 T1c), T2 in 109 patients (75 T2a and 34 T2b), and T3 in 35 patients. The median pretreatment PSA level was 10.8 ng/mL (range, 1.1–206) and the median Gleason score was 6 (range, 2–10). Clinical stage, pretreatment PSA level, and Gleason score were used to further categorize patients as defined above in the Materials and Methods section. Forty-seven patients (21%) were low risk, 60 (27%) were intermediate risk, and 115 (52%) were high risk. One hundred eighty-two patients received hormonal therapy before and/or during irradiation. The median follow-up after radiotherapy for surviving patients was 47 months (range, 0–99 months).

Table 1. Study Population Characteristics
Characteristics< 7560 cGy (n = 70) (%)7560 cGy (n = 152) (%)Overall (n = 222) (%)P valuea
  • cGy: Centigray; PSA: prostate-specific antigen.

  • a

    P value reflects the relation of each characteristic with dose selection.

  • b

    Unknown values are not included in the chi-square analysis.

  • c

    Likelihood ratio chi-square value.

Age (yrs)   0.710
 ≤6515 (21)36 (24)51 (23) 
 >6555 (79)116 (76)171 (77) 
Racebc   0.039
 White64 (91)118 (78)182 (82) 
 African American04 (3)4 (2) 
 Hispanic5 (7)21 (14)26 (12) 
 Asian03 (2)3 (1) 
 Unknown1 (1)6 (4)7 (3) 
Clinical T Classification   0.003
 T117 (24)61 (40)78 (35) 
  T1a01 (1)1 (0.5) 
  T1b2 (3)2 (1)4 (2) 
  T1c15 (21)58 (38)73 (33) 
 T246 (66)63 (41)109 (49) 
  T2a44 (63)31 (20)75 (34) 
  T2b2 (3)32 (21)34 (15) 
 T37 (10)28 (18)35 (16) 
Pretreatment PSAb   0.861
 ≤10 ng/mL33 (47)67 (44)100 (45) 
 10.01–19.99 ng/mL18 (26)44 (29)62 (28) 
 ≥20 ng/mL18 (26)41 (27)59 (27) 
 Unknown1 (1)01 (0.5) 
Gleason scoreb   0.217
 ≤641 (59)77 (51)118 (53) 
 719 (27)51 (34)70 (32) 
 ≥85 (7)21 (14)26 (12) 
 Unknown5 (7)3 (2)8 (4) 
Androgen deprivation   0.001
 Yes43 (61)139 (91)182 (82) 
 No27 (39)13 (9)40 (18) 
Median follow-up (mos)c5537470.574

The 2 and 5-year actuarial biochemical control rates for all patients were 83% and 77%, respectively The 2 and 5-year actuarial biochemical control rates were 96% and 89%, respectively, for low-risk patients, 84% and 77% for intermediate-risk patients, and 78% and 72% for high-risk patients (Table 2; Fig. 1). To assess the influence of the various prognostic features on biochemical disease recurrence-free survival rates, a multivariate analysis using logistic regression was conducted. Logistic regression identified lower dose (P = 0.006), higher risk group (P = 0.033), higher stage of disease (P = 0.045), and higher PSA level (P = 0.0001) as independent predictors of biochemical disease recurrence (Table 3). Gleason score, patient age, and use of hormonal therapy were not independent predictors.

Table 2. Actuarial Biochemical Control Rates
Risk group2-Yr rate (%)5-Yr rate (%)
Low9689
Intermediate8477
High7872
Overall8377
Figure 1.

Actuarial probability of biochemical survival years by risk group. There were 47 patients in the low-risk group, 56 patients in the medium-risk group, and 112 patients in the high-risk group. Seven patients were omitted because of unknown status. The number of patients at risk at 2 years was 39 in the low-risk group, 33 in the medium-risk group, and 70 in the high-risk group; the number at risk at 5 years was 9 in the low-risk group, 7 in the medium-risk group, and 18 in the high-risk group. A log-rank test comparison of curves was performed, with a P value of 0.072 for the low-risk versus the medium-risk group and a P value of 0.018 for the low-risk group versus the high-risk group=. Solid line: high-risk group; dotted line: medium-risk group; dashed line: low-risk group.

Table 3. Factors Affecting Biochemical Recurrence-Free Survival
PredictorOdds ratioProbability
  • PSA: prostate-specific antigen; RT: radiotherapy.

  • a

    Not significant at P ≤ 0.05.

  • b

    Lower values were associated with a higher probability of disease recurrence.

  • c

    Higher values were associated with a higher probability of disease recurrence.

Doseb0.380.006
Risk groupc1.690.033
Stagec1.280.045
PSA (maximum before RT)c1.020.001
Gleason1.270.06, NSa
Age0.980.37, NSa
Hormones1.040.93, NSa

The 5-year actuarial cause-specific survival rate was 95%. Clinical disease recurrence occurred in 13 patients at a median of 43 months (range, 23–99 months).

Within the first 3 months, only 8% and 1% of patients experienced Grade 2 and Grade 3 GI toxicity, respectively. After 3 months, only 7% and 2% of patients experienced Grade 2 and Grade 3 GI toxicity, respectively. No patients experienced Grade 4 GI toxicity. Also, 30% and 1% of patients experienced Grade 2 and Grade 3 GU toxicity, respectively, within the first 3 months, and only 9% had Grade 2 toxicity after 3 months. One patient experienced Grade 4 bladder necrosis. The distribution of acute and late bladder and rectal toxicity by dose received is shown in Table 4.

Table 4. Distribution of Patients by Acute and Late Bladder and Rectal Toxicity by Dose in cGy
Toxicity characteristics by dose (cGy)Grade 0 (%)Grade 1 (%)Grade 2 (%)Grade 3 (%)Grade 4 (%)No follow-up available (%)P valuea
  • Gu: genitourinary; GI: gastrointestinal.

  • a

    P value indicates the relation of dose to development of toxicity.

Acute GU      0.378
 <75603 (4)48 (69)19 (27)000 
 75608 (5)93 (61)48 (32)3 (2)00 
 Overall11 (5)141 (64)67 (30)3 (1)00 
Late GU      0.125
 <756035 (50)23 (33)6 (9)01 (1) 5 (7) 
 756053 (35)67 (44)14 (9)1 (1)017 (11) 
 Overall88 (40)90 (41)20 (9)1 (0.5)1 (0.5)22 (10) 
Acute GI      0.318
 < 756032 (46)30 (43)7 (10)1 (1)00 
 756080 (53)62 (41)10 (7)000 
 Overall112 (50)92 (41)17 (8)1 (0.5)00 
Late GI      0.016
 < 756030 (43)25 (36)7 (10)2 (3)0 6 (9) 
 756092 (61)28 (18)8 (5)2 (1)022 (14) 
 Overall122 (55)53 (24)15 (7)4 (2)028 (13) 

During treatment, 185 (83%) patients required some medication for symptom management, most commonly ibuprofen, which was prescribed to slightly more than one-half of the patients treated. Other commonly prescribed medications for symptom management included peripheral anti-alpha adrenergic agents, laxatives, anti-diarrheals, and topical corticosteroids. The most common GU symptom was urinary frequency, and common GI complaints included rectal pain, mild bleeding, and diarrhea.

A t test was used to compare average GU/GI toxicity of low (< 7560 cGy) and high-dose (7560 cGy) groups. Development of acute and late toxicity was not significantly correlated with higher radiotherapy dose. As shown in Table 4, development of late GI toxicity was statistically, significantly associated with lower radiotherapy dose (P = 0.016).

Of the 54 patients who were known to be sexually potent before irradiation, 33% (n = 18) had retained potency, 30% (n = 16) reported partial loss of sexual function, and 37% (n = 20) reported complete loss of sexual function at last follow-up. Chi-square analysis showed that loss of potency was correlated with higher radiotherapy dose (P = 0.013). Although there were few patients in each subcategory, no significant relations was found between loss of potency and hormones.

DISCUSSION

Multiple retrospective studies and one prospective randomized trial have now validated the superiority of three-dimensional conformal radiotherapy and dose escalation over conventional therapy in terms of biochemical control1, 3, 4, 7, 8, 12, 14, 17, 18, 21, 23, 29, 30 and toxicity1, 2, 5, 6, 9, 10, 22, 30–38 in the treatment of localized prostate carcinoma. In 1998, Zelefsky et al.4 reported on 743 patients treated with 3DCRT at Memorial Sloan Kettering Cancer Center, the institution on which we based our treatment approach. With a median follow-up of 3 years, the 5-year actuarial biochemical control rates for favorable (T1–2 disease, pretreatment PSA level ≤ 10.0 ng/mL, and Gleason score ≤ 6), intermediate (1 of the previous prognostic factors with a higher value), and unfavorable (≥ 2 factors with a higher value) groups were 85%, 65%, and 35%, respectively. These results were updated in 2000 to 1100 patients with a median follow-up of 52 months. At that time, the authors reported 5-year actuarial biochemical control rates of 85%, 59%, and 39% for favorable, intermediate, and unfavorable risk groups, respectively.5 In comparison, our current study demonstrates similar rates of biochemical control for low-risk patients (89%) and even somewhat higher rates of biochemical control for intermediate and high-risk patients (79% and 72%, respectively). Direct comparison of biochemical control rates between studies is not strictly valid given the number and magnitude of uncontrolled differences including differences in patient selection, risk group criteria, treatment techniques, specification of target volumes, dosing regimens, and length of follow-up. Nonetheless, a tabulation of recent series can provide important gross benchmarks for expected treatment outcomes (Table 5). In terms of biochemical control rates (both overall and stratified by risk group), our results compare favorably with other published series.

Table 5. Biochemical Control Rates as Reported in Recent Studies Regarding 3DCRT in Prostate Carcinoma
CharacteristicsMedian follow-up (mos)No. of patientsLow risk (%)Intermediate risk (%)High risk (%)Overall risk (%)
  1. 3DCRT: three-dimensional conformal radiotherapy.

Current study4722289797278
Zelefsky et al., 1998(4)36743856535 
Zelefsky et al., 2000(5)521100855939 
Lyons et al., 2000(15)45738   58
Kupelian et al., 2000(14)331041   61
Fiveash et al., 2000(23)36180  62.5 
Roach et al., 1996(12)2450  23 
Hanks et al., 2002(20)110229   55
Magrini et al., 1998(30)Not given208   60
Seung et al., 1998(13)34187   75
Fukunaga-Johnson et al., 1997(22)3670775 37 
Pinover et al., 2000(19)3648885 81 

Previous studies have also identified independent disease and treatment predictors of biochemical disease recurrence. Radiotherapy dose has been found consistently to affect rates of biochemical control,1, 3, 7, 8, 12, 14, 17, 18, 20, 21, 23, 29, 30 as is the case with the current study. Two studies have demonstrated the significance of the Gleason score in terms of biochemical outcome.12, 23 However, we were not able to show that association in the current study. This is possibly due to the relatively few patients with poorly differentiated lesions. We found that risk group, tumor stage, and pretreatment PSA level were significant determinants of biochemical disease recurrence, supporting earlier reports. It is noteworthy that the use of hormones did not play a significant role in biochemical control. Use of androgen deprivation along with radiotherapy proved biochemical disease recurrence-free survival, cause-specific survival, and overall survival in randomized clinical trials for patients with locally advanced or high-risk disease.39–41 As is common in community practices such as ours, patients often begin androgen deprivation therapy before consultation with a radiation therapist and according to nonstandardized selection criteria. Thus, > 80% of our patients were treated with hormones, which may have masked any potential benefit that may have been seen with more strict criteria for use of hormones.

Several randomized and nonrandomized controlled trials have demonstrated either a reduction in acute and late toxicity with 3DCRT or no significant difference in toxicity despite higher doses in the conformal arm.1, 2, 5, 6, 9, 10, 22, 30–38 Again, although direct comparison of toxicity outcomes between studies is not strictly valid given the different parameters from study to study, Table 6 provides a general sense of the range of expected acute and chronic GI and GU toxicity in studies that used similar total doses and the same grading scheme (RTOG toxicity scales). The frequency and degree of toxicities in the current study are at least grossly comparable to those achieved in other institutions.

Table 6. Toxicity Outcomes as Reported in Recent Studies on 3DCRT in Prostate Carcinoma
OutcomesGIGU
Grade 2 (%)Grade 3 (%)Grade 4 (%)Grade 2 (%)Grade 3 (%)Grade 4 (%)
  1. 3DCRT: three-dimensional conformal radiotherapy; GI: gastrointestinal; GU: genitourinary.

Acute      
 Current study80.503010
 Bergstrom et al., 1998(31)13002500
 Kupelian et al., 2000(14)18001910
 Leibel et al., 1994(6) 00 < 10
 Zelefsky et al., 1999(2)1700370.70
 Nuyttens et al., 2002(38) 00   
Chronic      
 Current study72090.50.5
 Kupelian et al., 2000(14)1230   
 Leibel et al., 1994(6)  < 1 < 10
 Bey et al., 2000(33) 00 70
 Michalski et al., 2000(34) 00 < 10
 Hanks et al., 1998(1) < 1< 1 44
 Zelefsky et al., 1999(2)110.7501030
 Fukunaga-Johnson et al., 199722 33 10
 Huang et al., 2002(11)17.85.50   
 Dearnaley et al., 1999(37) 00 < 50

The presumed benefit of 3DCRT lies in its ability to better conform radiotherapy to the target volume, thus allowing for dose escalation and improved tumor control with limited morbidity. This expected benefit occurred in our experience with biochemical control rates and toxicity outcomes that generally compare favorably with previously published experiences, primarily from large academic institutions. Given the recent M. D. Anderson Cancer Center randomized study demonstrating the superiority of dose escalation with conformal radiotherapy3 and RTOG 94-0634 demonstrating the safety of this method at doses ≤ 79 Gy, we believe the bar has been set for such technology to be applied routinely in community-based practice. Review of the treatment volumes contoured by expert academic radiation oncologists at the previous three ASTRO 3DIMRT practicums have demonstrated significant variation in CTV delineation using standard test cases (unpublished data). We have noted a significant variation in volume delineation when we first used our 3DCRT technique in other institutions within our medical group. We have developed a quality control program using “screen shots” exported from the treatment planning systems at our institutions to a Health Insurance Portability and Accountability Agt (HIPPA) compliant web server so that 3D target volumes can be contoured uniformly and peer reviewed by our other physicians.

With the excellent results of the current study and the evolution of technology for limiting doses to normal tissues and dose escalating at the disease, we have begun using intensity-modulated radiotherapy (IMRT) with daily B-mode ultrasound localization to doses of 78–80 Gy. As the newer technology of IMRT and treatment techniques afforded by this method diffuse into community-based practices, it is incumbent upon us to demonstrate the reproducibility of good results in the community setting. The rapid pace of technology acquisition and acceptance demands that treatment results be benchmarked in comparison to early published results. Otherwise, the majority of patients treated for malignant disease in our mixed health care delivery system are at risk simultaneously for the poorer outcomes of decreased local control and increased complication. This report of our current community-based practice is an attempt to provide such data. We encourage other community-based practices to perform similar analyses in this era of dose escalation, increased conformality of treatment, and narrower CTV-PTV expansions.

Ancillary