- Email: [email protected]
Optimizing patient selection for dose escalation techniques using the prostate-specific antigen level, biopsy gleason score, and clinical T-stage
Int. J. Radiation Oncology Biol. Phys., Vol. 45, No. 5, pp. 1227–1233, 1999 Copyright © 1999 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/99/$–see front matter
PII S0360-3016(99)00303-X
CLINICAL INVESTIGATION
Prostate
OPTIMIZING PATIENT SELECTION FOR DOSE ESCALATION TECHNIQUES USING THE PROSTATE-SPECIFIC ANTIGEN LEVEL, BIOPSY GLEASON SCORE, AND CLINICAL T-STAGE ANTHONY V. D’AMICO, M.D., PH.D.,* RICHARD WHITTINGTON, M.D.,† S. BRUCE MALKOWICZ, M.D.,‡ DELRAY SCHULTZ, PH.D.,§ ANDREW A. RENSHAW, M.D.,㛳 JOHN E. TOMASZEWSKI, M.D.,¶ JEROME P. RICHIE, M.D.,# AND ALAN WEIN, M.D.‡ *Joint Center for Radiation Therapy, Harvard Medical School, Boston, MA; Departments of †Radiation Oncology, ‡Urology, and Pathology, Hospital of the University of Pennsylvania, Philadelphia, PA; §Department of Mathematics, University of Millersville, Millersville, PA; and Departments of 㛳Pathology and #Urology, Brigham and Women’s Hospital, Boston, MA.
¶
Purpose: Ideal candidates for 3D dose escalation conformal radiation or external beam ⴙ implant therapy are identified on the basis of the prostate-specific antigen (PSA) level, biopsy Gleason score, and the 1992 American Joint Commission Cancer (AJCC) clinical T-stage. Methods and Materials: The pathologic findings of 1742 men with clinical stage T1c,2 prostate cancer managed with a radical prostatectomy (RP) between 1990 and 1998 were subjected to a logistic regression multivariable analysis. The endpoints examined included pathologic organ– confined (OC), specimen-confined (SC), and margin (M) or seminal vesicle (SV) positive disease. SC disease was defined as extracapsular extension (ECE) with a negative surgical margin. The clinical factors tested included PSA level, biopsy Gleason score, and the 1992 AJCC clinical T-stage. PSA failure–free (bNED) survival was calculated according to the method of Kaplan and Meier. Results: Significant negative predictors of pathologic OC– disease or positive predictors of Mⴙ or SVⴙ disease included a PSA > 10 ng/ml (p < 0.0001), biopsy Gleason score >7 (p < 0.0004), and > T2b disease (p < 0.03). Only biopsy Gleason score 7 (p ⴝ 0.0006) and PSA 10 –15 ng/ml (p ⴝ 0.04) were significant predictors of SC disease. The estimates of 5-year bNED survival were 80%, 62%, and 35% (p < 0.0001) for patients having a low, intermediate, or high likelihood of having Mⴙ or SVⴙ disease respectively. Conclusions: Patients most likely to derive a survival benefit from the improved local control possible using dose escalation techniques were those who had both a low risk of having occult micrometastatic disease (<25% Mⴙ or SVⴙ) and a reasonable likelihood of remaining disease-free after RP (>50% 5-year bNED). These patients included those having T1c, 2a, PSA > 10 –15 ng/ml, and biopsy Gleason <6 or T1c, 2a, 2b, PSA < 10 ng/ml, and biopsy Gleason < 7 prostate cancer. © 1999 Elsevier Science Inc. Prostate cancer, 3D conformal dose escalation, Implant boost, Patient selection, Prostate-specific antigen.
INTRODUCTION
Therefore, selecting the patient likely to derive a survival benefit from the use of dose escalation radiotherapy techniques would have two requirements. First, the probability of seminal vesicle invasion, margin-positive or lymphnode–positive disease would need to be low in order to minimize the risk of occult micrometastatic disease. Second, there would need to be a high likelihood that radical local therapy could ablate the entire volume of cancer. The likelihood of success in this regard may be approximated from the prostate-specific antigen (PSA) failure–free (bNED) survival after RP. Identifying these patients prior to therapy, on the basis of the readily available pretreatment prognostic factors was the goal of this study.
Data from the surgical literature (1, 2) have suggested that the cancer-specific survival after radical prostatectomy (RP) decreases as one advances from pathologic organ– confined to specimen-confined (extracapsular extension [ECE] with negative surgical margins) to positive-margin, seminal vesicle–positive, and finally lymph node–positive disease. These decrements in cancer-specific survival could be due to local failure that in time metastasizes and becomes a distant failure. However, a significant proportion of these deaths are likely due to the increasing incidence of occult distant micrometastatic disease present in men with a more advanced pathologic stage at the time of the RP.
Reprint requests to: Anthony V. D’Amico, M.D., Ph.D., Joint Center for Radiation Therapy, Harvard Medical School, 330 Brookline Avenue–5th floor, Boston, MA 02215. E-mail: [email protected]
Accepted for publication 22 July 1999.
1227
1228
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 45, Number 5, 1999
Table 1. Clinical pretreatment and postoperative pathologic characteristics of the 1742 surgically managed patients studied using the logistic regression multivariable analyses Clinical characteristic
Number
Pathologic OC
Pathologic SC
Pathologic M⫹ or SV⫹
AJCC clinical T1c AJCC clinical T2a AJCC clinical T2b AJCC clinical T2c Biopsy Gleason 2–4 Biopsy Gleason 5–6 Biopsy Gleason 7 Biopsy Gleason 8–10 PSA ⱕ 4 ng/ml PSA ⬎ 4–10 ng/ml PSA ⬎ 10–20 ng/ml PSA ⬎ 20 ng/ml
688 653 214 187 280 999 343 120 184 1044 367 147
529 (77%) 455 (70%) 118 (55%) 89 (48%) 224 (80%) 727 (73%) 185 (54%) 55 (46%) 151 (82%) 786 (75%) 206 (56%) 48 (33%)
65 (9%) 85 (13%) 30 (14%) 23 (12%) 22 (8%) 109 (11%) 59 (17%) 13 (11%) 15 (8%) 113 (11%) 53 (15%) 22 (15%)
94 (14%) 113 (17%) 66 (31%) 75 (40%) 34 (12%) 163 (16%) 99 (29%) 52 (43%) 18 (10%) 145 (14%) 108 (29%) 77 (52%)
Abbreviations: AJCC ⫽ 1992 American Joint Commission on Cancer Staging System; PSA ⫽ prostate-specific antigen; SC ⫽ specimen-confined; OC ⫽ organ-confined; SV ⫽ seminal vesicle; M ⫽ margin; ⫹ ⫽ Positive.
METHODS AND MATERIALS Patient population Seventeen hundred and forty-two men treated with a RP and bilateral pelvic lymph node dissection at the Hospital of the University of Pennsylvania (HUP) or the Brigham and Women’s (B&W) Hospital between 1990 and 1998 who had PSA detected or clinically palpable prostate cancer comprised the study population. Table 1 lists the preoperative clinical and postoperative pathological characteristics of the 1742 study patients. Preoperative staging In all cases, staging evaluation included a history and physical examination including a digital rectal exam (DRE), serum PSA, computed tomography (CT) of the pelvis or an endorectal and pelvic coil magnetic resonance imaging (MRI) scan of the prostate and pelvis, bone scan, and a transrectal ultrasound guided (TRUS) needle biopsy of the prostate with Gleason score histologic grading (3). The prostate biopsy was performed using an 18-gauge Tru-Cut needle (Travenol Laboratories, Deerfield, IL) via a transrectal approach. The clinical stage was obtained from the DRE findings using the current 1992 AJCC staging system (4). Radiologic and biopsy information were not used to determine clinical stage. The PSA was obtained on an ambulatory basis prior to radiologic studies and biopsy. All PSA measurements (5) were made using the Hybritech, Tosoh, or Abbot assays. Treatment and pathologic processing A referee genitourinary pathologist reviewed the diagnostic biopsy specimens for all patients undergoing surgery at the Hospital of the University of Pennsylvania (JET) or the Brigham and Women’s Hospital (AAR). Surgical treatment consisted of a radical retropubic prostatectomy and bilateral pelvic lymph node sampling. If the intraoperative frozen sections of any sampled lymph node were positive for carcinoma, the radical prostatectomy was aborted. Prosta-
tectomy specimens were weighed, measured, inked over the entire surface, and fixed in 10% buffered formalin. Both the apical and basal margins were amputated to a thickness of 5 mm and sectioned parasagitally in a direction perpendicular to the initial transverse incision at 3-mm intervals. The base of the seminal vesicles and the basal cross section were submitted separately for microscopic analysis. The prostate was sectioned perpendicular to the long axis (apical– basal) of the gland along the posterior/rectal surface at 5-mm intervals, with most specimens requiring 4 to 7 cross sections to be entirely sectioned. For each cross section, a single section each of the right and left posterior region was submitted. Finally, a single section from the mid anterior prostate was submitted for microscopic evaluation. Evidence of extraprostatic disease including seminal vesicle invasion (SVI), ECE, and/or positive surgical margins (PSM) were recorded. Disease extending into but not through the prostatic capsule was considered negative for ECE. Follow-up The median follow-up of the 960 and 782 surgically managed patients at HUP or the B&W Hospital was 46 months (8 –100 months) and 42 months (12–96 months) respectively. The patients were seen 1 month postoperatively, then at 3-month intervals for 2 years, every 6 months for 5 years, and annually thereafter. At each follow-up a serum PSA was obtained prior to performing the DRE. All pretreatment PSA values were obtained within 1 month of the date of surgery. No patient was lost to follow-up. Statistical analyses A logistic regression multivariable analysis (6) evaluating the ability of the preoperative PSA, biopsy Gleason score, and the 1992 AJCC clinical T-stage to predict pathologic organ– confined, specimen-confined (ECE with a negative surgical margin), and more advanced disease (margin-positive and/or seminal vesicle–positive) was performed. The
Prostate cancer patient selection for dose escalation
PSA, biopsy Gleason score, and 1992 AJCC stages were first treated as continuous variables and then in a separate analysis treated as categorical variables. In this latter analysis, PSA was grouped as ⬎0 – 4, ⬎4 –10, ⬎10 –15, ⬎15– 20, and ⬎20 ng/ml. Biopsy Gleason score was grouped as 2– 4, 5– 6, 7, and 8 –10. Finally the 1992 AJCC clinical T-stage was grouped as T1c, T2a, T2b, and T2c. Baseline groups for the categorical variables to which other groups were compared were PSA ⬎ 0 – 4 ng/ml, biopsy Gleason 2– 4, and 1992 AJCC clinical stage T1c, 2a. PSA failure was scored when two consecutive detectable PSA values were obtained postoperatively after an undetectable value. The time of PSA failure was then recorded at the time of the first detectable PSA. Time zero was defined as the day of surgery. If a PSA never became undetectable postoperatively, then PSA failure was defined to be at time equal to zero. For illustrative purposes, PSA failure–free survival (bNED) as a function of the pretreatment PSA, biopsy Gleason score, and 1992 AJCC clinical T-stage was estimated using an actuarial calculation according to the method of Kaplan and Meier (7) and graphically displayed. Patients found to have positive pelvic lymph nodes at the time of frozen section or whose pelvic lymph nodes were negative at the time of frozen section but positive on the final sections were started on androgen suppression therapy. Therefore, while these men (n ⫽ 12) were excluded from the Kaplan-Meier analysis of PSA survival, they were included in the margin-positive or SV category for the purpose of the logistic regression analysis.
RESULTS Clinical predictors of pathologic stage The significant negative predictors of pathologic organ– confined disease were analogous to the positive predictors of positive margins or seminal vesicle invasion. These clinical predictors included a PSA ⬎ 10 ng/ml ( p ⬍ 0.0001), biopsy Gleason score ⱖ 7 ( p ⱕ 0.0004), and ⱖ T2b disease ( p ⱕ 0.03). Only biopsy Gleason score 7 ( p ⫽ 0.0006) and PSA ⬎ 10 –15 ng/ml ( p ⫽ 0.04) were significant predictors of specimen-confined disease. Table 2 lists the complete set of p-values for the logistic regression analyses that evaluated the ability of the PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage to predict the various pathologic endpoints. Listed in Table 2 are the results when the clinical indicators were examined as either continuous or categorical variables. Table 3 rank orders the combinations of PSA, biopsy Gleason score, and 1992 AJCC clinical T-stage beginning with those combinations in Groups 1–5 that represent the patients who are most likely to have pathologic organ– confined disease and remain PSA failure–free (bNED) at 5 years after RP. These patients are followed by patients in groups 6 –9 who were more likely to have margin-positive disease or seminal vesicle invasion and least likely to remain PSA failure–free at 5 years following RP.
● A. V. D’AMICO et al.
1229
Table 2. P-values derived from the logistic regression multivariable analysis for each of the pathologic endpoints Clinical characteristic PSA (continuous) Biopsy Gleason score (continuous) 1992 AJCC clinical T-stage (continuous) PSA 0–4 ng/ml PSA ⬎ 4–10 ng/ml PSA ⬎ 10–15 ng/ml PSA ⬎ 15–20 ng/ml PSA ⬎ 20 ng/ml Biopsy Gleason 2–4 Biopsy Gleason 5–6 Biopsy Gleason 7 Biopsy Gleason 8–10 1992 AJCC T1c or T2a 1992 AJCC T2b 1992 AJCC T2c
Organconfined*
Specimen confined
Margin or SV ⫹
⬍ 0.0001
0.004
⬍ 0.0001
⬍ 0.0001
0.006
⬍ 0.0001
⬍ 0.0001 —† 0.11 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001 — 0.06 ⬍ 0.0001 0.0004 — 0.007 0.03
0.33 — 0.20 0.04 0.16 0.26 — 0.22 0.0006 0.84 — 0.31 0.97
0.0001 — 0.27 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001 — 0.28 0.0001 0.0001 — 0.01 0.01
* All significant p-values for this pathologic group had negative coefficients meaning that the likelihood of having pathologic organ– confined disease decreases significantly with an increasing value of the clinical predictor. † Denotes a baseline group for the categorical variable.
bNED Survival stratified by clinical and pathologic factors Figure 1 illustrates the decrement in the estimated 5-year bNED survival with increasing pathologic stage. Specifically, the estimated 5-year bNED rates were 77%, 53%, and 20% for patients with organ-confined, specimen-confined, or at least margin-positive disease. All pairwise p-values for comparisons among these groups of patients were ⬍0.0001. Figure 2 illustrates a similar trend for decreasing estimates of bNED survival as one moves from a group of patients whose clinical pretreatment factors as shown in Table 3 are most often associated with pathologic organ– confined disease (Group 1) to those most often associated with at least margin-positive disease (Groups 6 –9). Specifically, the 5-year bNED rates for patients in Groups 1, 2–5, and 6 –9 are 80%, 62%, and 35% respectively. The pairwise p-values for all comparisons in Fig. 2 were also ⬍0.0001. The justification for the groupings in Figure 2 is shown in Fig. 3 where the estimates of the 5-year bNED survival after RP are 80% for Group 1, 68%, 53%, 68%, and 58% ( p ⬎ 0.05) for Groups 2, 3, 4, and 5 respectively and 37%, 37%, 39%, and 33% ( p ⬎ 0.05) for Groups 6, 7, 8, and 9 respectively. An additional subgroup analysis for each grouping (1 vs. 2–5 vs 6 –9) was performed stratified by pathologic stage to assess PSA outcome in these groups. Figures 4 – 6 depict the bNED survival in these groups for men with organ-confined, specimen-confined, and margin- or SV-positive disease respectively. The 4-year bNED rates for patients in groups 1 vs. 2–5 vs. 6 –9 were 87%, 75%, 70% for men with pathologic organ– confined disease. Similarly these values
1230
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 45, Number 5, 1999
Table 3. Rank ordering of the pretreatment clinical characteristic groups by the frequency of the pathologic outcomes and 5-year PSA failure–free (bNED) survival Group
PSA
b Gleas
T-stage
N
OC
SC
M⫹ or SV⫹*
5-yr bNED
1 2 3 4 5 6 7 8 9
ⱕ10 ⱕ10 ⬎10–15 ⱕ10 ⱕ10 ⬎10–15 ⬎10–15 ⬎10–15
ⱕ6 7 ⱕ6 ⱕ6 7 ⱕ6 7 7 ⬎15 or ⱖ8 or 2c
1c, 2a 1c, 2a 1c, 2a 2b 2b 2b 1c, 2a 2b
801 179 144 87 27 22 34 10 438
81% 67% 70% 76% 52% 39% 35% 40% 48%
9% 20% 13% 7% 26% 23% 24% 20% 14%
10% 13% 17% 17% 22% 38% 41% 40% 38%
80% 68% 53% 68% 58% 37% 37% 39% 33%
* Twelve patients had lymph node ⫹ disease and were included in the M⫹ or SV⫹ category. Abbreviations: b Gleas ⫽ Biopsy Gleason score; T-stage ⫽ 1992 AJCC clinical T-stage; N ⫽ number; OC ⫽ organ-confined; SC ⫽ specimen-confined; M ⫽ margin; SV ⫽ seminal vesicle; ⫹ ⫽ positive.
were 75%, 62%, 37% and 58%, 42%, 15% for men with specimen-confined and margin- or SV-positive disease respectively.
Retrospective comparisons of patients managed with 3D dose escalation conformal external beam RT (8 –10) or external beam ⫹ interstitial implant boost (11, 12) versus conventional-dose external beam RT have reported statistically significant improvements in bNED survival rates. One report from Hanks and colleagues (8) has suggested an improvement in overall survival for men treated with 3D dose escalation RT versus historical controls treated with external beam RT to conventional dose. While these findings await prospective validation, identifying patients today who are most likely to derive a survival benefit from the use of dose escalation techniques is a clinically relevant issue.
Selecting the ideal patient for dose escalation radiation therapy would have two requirements. First, the probability of seminal vesicle invasion, margin-positive, or lymph node–positive disease would need to be low in order to minimize the risk of occult micrometastatic disease. In this study as in others (1, 2) bNED survival after RP was found to be significantly less for patients with specimen-confined, margin-positive disease, or seminal vesicle invasion when compared to pathologic organ– confined and margin-negative disease (Fig. 1). Second, in order to ensure that local only therapy is sufficient for cure, 5-year estimates of bNED survival after RP (Table 3) could be used to approximate success. Using bNED survival after RP as the surrogate for measuring the likelihood of success is reasonable considering that the margins obtained in a RP specimen are less than or equal to the margins used when utilizing higher RT doses either with 3D conformal dose escalation or external beam RT plus an interstitial radiation therapy boost. Specifically,
Fig. 1. PSA (bNED) survival stratified by pathologic organ– confined (OC), specimen-confined (SC), or at margin (M) or seminal vesicle (SV)-positive disease for the 1742 study patients. All pairwise p-values ⬍ 0.0001.
Fig. 2. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3. All pairwise p-values ⬍ 0.0001.
DISCUSSION
Prostate cancer patient selection for dose escalation
● A. V. D’AMICO et al.
1231
Fig. 3. PSA (bNED) survival stratified by specific groupings of the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3.
Fig. 5. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3 for patients with pathologic specimen– confined (SC) disease.
a negative margin in a RP specimen can range from less than 1 mm to as much as 5 mm. This is similar to the smaller RT field margins used at the prostatic–rectal interface (⬃5 mm) when escalating RT dose (9) beyond the conventionaldose range. An additional complexity of smaller margins however, is that the intrinsic prostate motion has been reported to be on the order of 2–3 mm in the anterior– posterior direction (13) which can also impact on the volume of the prostate receiving the prescription dose. Therefore, requiring both that the estimates of 5-year bNED survival after RP are reasonable (⬎50%) and that the patient has a low risk of occult micrometastatic disease could translate into an improvement in survival.
Defining the optimal candidates for dose escalation RT Evaluating the results of Table 3 and Figure 3 provides the basis on which to define 3 distinct patient subgroups. These 3 groups can be defined using the 5-year bNED survival rates and pathologic characteristics found upon pathologic evaluation of the RP specimen. Specifically, patients in Groups 6 –9 who had the lowest 5-year bNED rates (33–39%) also had the highest rates of margin- or SV-positive disease (38 – 41%). These patients would be expected to have the highest risk of occult micrometastatic disease and therefore are least likely to derive a survival benefit from attempts at improving local therapy. Evaluating the impact of systemic therapy and improved local therapies in patient Groups 6 –9 would be reasonable. Such
Fig. 4. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3 for patients with pathologic organ– confined (OC) disease.
Fig. 6. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3 for patients with marginpositive or seminal vesicle–positive (M or SV ⫹) disease.
1232
I. J. Radiation Oncology
●
Biology
●
Physics
an endeavor is currently being performed by the Radiation Therapy Oncology Group in protocol 9408. RTOG 9408 is a Phase III trial evaluating the impact on survival of adding 4 months of total androgen suppression therapy to conventional dose external beam RT. It is interesting to note, however, that of all men in Groups 6 –9, 58% remained PSA failure–free 4 years following RP if they had organ-confined disease as noted in Fig. 4. This cohort of men with organ-confined disease comprised 47% of this high-risk subgroup. Therefore, attempts at improving local therapy through dose escalation in the high-risk patients is likely to improve outcome beyond the improvement noted when androgen suppression therapy is added to conventional-dose external beam radiation therapy (14). The second patient subgroup is comprised of the men in Group 1. These patients had the highest 5-year bNED rate (81%) and the lowest rate of margin- or SV-positive disease (10%). Therefore, these patients would be expected to have the lowest rate of occult micrometastatic disease as well as the smallest local tumor burden. Recently, patients in this subgroup have been suggested by Pinover and colleagues (10) to derive a benefit in bNED survival through the use of 3D dose escalation RT beyond 71.5 Gy. Patients comprising Groups 2–5 represent the final patient subgroup. The 5-year bNED survival of these men was intermediate between the patients in the most favorable (Group 1) and least favorable (Groups 6 –9) groups ranging from 53– 68%. Similarly the finding of margin- or SVpositive disease was also intermediate between the most and least favorable groups, being 13–22%. While some men in this group will have occult micrometastatic disease and will have failed RP on that basis, a significant fraction could
Volume 45, Number 5, 1999
have had a larger volume of local disease which may have been inadequately addressed by the limited margins that can be obtained using RP. In that case, these men would be the optimal candidates for investigating the impact that escalating the RT dose has on survival. In particular, as noted in Fig. 5, 75% and 62% of patients in Groups 2–5 remained PSA failure–free 4 years following RP if they had organconfined (n ⫽ 299) or specimen-confined (n ⫽ 61) disease respectively as noted in Fig. 5. These 360 patients comprised the vast majority (83%) of the whole cohort (n ⫽ 433) providing further evidence that dose escalation in this subgroup may improve outcome in the majority of these patients. Previous retrospective comparisons (15, 16) have reported comparable PSA failure–free survival for the patient groups analyzed in this study who were managed with either a RP or conventional-dose external beam RT. Therefore, it is reasonable to hypothesize that higher RT doses in the select patients whose risk of occult micrometastatic disease is low (Groups 1–5) may improve survival beyond that obtained using RP or conventional-dose RT. CONCLUSION Patients most likely to derive a survival benefit from the improved local control possible using dose escalation techniques in this study were those who had both a low risk of having occult micrometastatic disease (⬍25% M⫹ or SV⫹) and a reasonable likelihood of remaining disease-free after RP (⬎50% 5-year bNED). These patients included those having T1c, 2a, PSA ⬎ 10 –15 ng/ml, and biopsy Gleason ⱕ6 or T1c, 2a, 2b, PSA ⱕ 10 ng/ml, and biopsy Gleason ⱕ7 prostate cancer.
REFERENCES 1. Iselin CE, Robertson JE, Paulson DF. Radical perineal prostatectomy: Oncological outcome during a 20-year period. J Urol 1999;161:163–168. 2. Walsh PC, Partin AW, Epstein JI. Cancer control and quality of life following anatomical radical retropubic prostatectomy: Results at 10 years. J Urol 1994;152:1831–1836. 3. Gleason DF and the Veterans Administration Cooperative Urological Research Group. Histologic grading and staging of prostatic carcinoma. In: Tannenbaum M, editor. Urologic pathology. Philadelphia: Lea & Febiger; 1977. p. 171–187. 4. Beahrs OH, Henson DE, Hutter RVP. American Joint Committee on Cancer, Manual for staging cancer, 4th ed. Philadelphia: J.P. Lippincott; 1992. 5. Oesterling JE, Jacobsen SJ, Klee GG, et al. Free, complexed and total serum prostate specific antigen: The establishment of appropriate reference ranges for their concentrations and ratios. J Urol 1995;154:1090 –1095. 6. Simultaneous inferences and other topics in regression analysis-1. In: Neter J, Wasserman W, Kutner M, editors. Applied linear regression models, 1st ed. Homewood, IL: Richard D. Irwin, Inc; 1983. p. 150 –153. 7. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Amer Stat Assoc 1958;53:457–500. 8. Hanks GE, Hanlon AL, Horwitz E, Pinover W, Schultheiss T. A relationship of local prostate dose and subsequent metasta-
9.
10.
11.
12. 13.
sis with 3D CRT in prostate cancer (Abstract 38). Int J Radiat Oncol Biol Phys 1998;42S:143. Fiveash JB, Hanks GE, Roach M III, Wang S, Vigneault E, McLaughlin W, Sandler HM. 3D Conformal radiation therapy (3DCRT) for high grade prostate cancer: A multi-institutional review (Abstract 37). Int J Radiat Oncol Biol Phys 1998;42S: 143. Pinover W, Hanlon AL, Horwitz EM, Hanks GE. Defining the appropriate dose for prostate cancer patients with PSA ⬍ 10 ng/ml (Abstract 36). Int J Radiat Oncol Biol Phys 1998;42S: 142. Ragde H, Eigamal AA, Snow PB, Brandt J, Bartolucci AA, Nadir BS, Korb LJ. Ten-year disease free survival after transperineal sonography-guided iodine-125 brachytherapy with or without 45-Gray external beam irradiation in the treatment of patients with clinically localized, low to high Gleason grade prostate carcinoma. Cancer 1998;83:989 –1001. Critz FA, Levinson K, Williams WH, Holladay DA. Prostatespecific antigen nadir: The optimum level after irradiation for prostate cancer. J Clin Oncol 1996;14:2893–2900. Beard CJ, Kijewki P, Bussiere M, Gelman R, Gladstone D, Shaffer K, Plunkett M, Costello P, Coleman CN. Analysis of prostate and seminal vesicle motion: Implications for treatment planning. Int J Radiat Oncol Biol Phys 1996;34:451– 458.
Prostate cancer patient selection for dose escalation
14. Bolla M, Gonsalez D, Warde P, Dubois JB, Mirimanoff R, Storme G, et al. Improved survival in patients with locally advanced prostate cancer treated with radiotherapy and goserelin. N Eng J Med 1997;337:295–300. 15. D’Amico AV, Whittington R, Malkowicz SB, Schultz D, Tomaszewski JE, Kaplan I, Beard C, Wein A. Biochemical outcome after radical prostatectomy, external beam radia-
● A. V. D’AMICO et al.
1233
tion therapy, or interstitial radiation therapy for clinically localized prostate cancer. J Am Med Assoc 1998;280:969 – 974. 16. Keyser D, Kupelian PA, Zippe CD, Levin HS, Klein EA. Stage T1-2 prostate cancer with pre-treatment prostate-specific antigen level ⱕ 10 ng/ml: Radiation therapy or surgery? Int J Radiat Oncol Biol Phys 1997;38:723–729.
PII S0360-3016(99)00303-X
CLINICAL INVESTIGATION
Prostate
OPTIMIZING PATIENT SELECTION FOR DOSE ESCALATION TECHNIQUES USING THE PROSTATE-SPECIFIC ANTIGEN LEVEL, BIOPSY GLEASON SCORE, AND CLINICAL T-STAGE ANTHONY V. D’AMICO, M.D., PH.D.,* RICHARD WHITTINGTON, M.D.,† S. BRUCE MALKOWICZ, M.D.,‡ DELRAY SCHULTZ, PH.D.,§ ANDREW A. RENSHAW, M.D.,㛳 JOHN E. TOMASZEWSKI, M.D.,¶ JEROME P. RICHIE, M.D.,# AND ALAN WEIN, M.D.‡ *Joint Center for Radiation Therapy, Harvard Medical School, Boston, MA; Departments of †Radiation Oncology, ‡Urology, and Pathology, Hospital of the University of Pennsylvania, Philadelphia, PA; §Department of Mathematics, University of Millersville, Millersville, PA; and Departments of 㛳Pathology and #Urology, Brigham and Women’s Hospital, Boston, MA.
¶
Purpose: Ideal candidates for 3D dose escalation conformal radiation or external beam ⴙ implant therapy are identified on the basis of the prostate-specific antigen (PSA) level, biopsy Gleason score, and the 1992 American Joint Commission Cancer (AJCC) clinical T-stage. Methods and Materials: The pathologic findings of 1742 men with clinical stage T1c,2 prostate cancer managed with a radical prostatectomy (RP) between 1990 and 1998 were subjected to a logistic regression multivariable analysis. The endpoints examined included pathologic organ– confined (OC), specimen-confined (SC), and margin (M) or seminal vesicle (SV) positive disease. SC disease was defined as extracapsular extension (ECE) with a negative surgical margin. The clinical factors tested included PSA level, biopsy Gleason score, and the 1992 AJCC clinical T-stage. PSA failure–free (bNED) survival was calculated according to the method of Kaplan and Meier. Results: Significant negative predictors of pathologic OC– disease or positive predictors of Mⴙ or SVⴙ disease included a PSA > 10 ng/ml (p < 0.0001), biopsy Gleason score >7 (p < 0.0004), and > T2b disease (p < 0.03). Only biopsy Gleason score 7 (p ⴝ 0.0006) and PSA 10 –15 ng/ml (p ⴝ 0.04) were significant predictors of SC disease. The estimates of 5-year bNED survival were 80%, 62%, and 35% (p < 0.0001) for patients having a low, intermediate, or high likelihood of having Mⴙ or SVⴙ disease respectively. Conclusions: Patients most likely to derive a survival benefit from the improved local control possible using dose escalation techniques were those who had both a low risk of having occult micrometastatic disease (<25% Mⴙ or SVⴙ) and a reasonable likelihood of remaining disease-free after RP (>50% 5-year bNED). These patients included those having T1c, 2a, PSA > 10 –15 ng/ml, and biopsy Gleason <6 or T1c, 2a, 2b, PSA < 10 ng/ml, and biopsy Gleason < 7 prostate cancer. © 1999 Elsevier Science Inc. Prostate cancer, 3D conformal dose escalation, Implant boost, Patient selection, Prostate-specific antigen.
INTRODUCTION
Therefore, selecting the patient likely to derive a survival benefit from the use of dose escalation radiotherapy techniques would have two requirements. First, the probability of seminal vesicle invasion, margin-positive or lymphnode–positive disease would need to be low in order to minimize the risk of occult micrometastatic disease. Second, there would need to be a high likelihood that radical local therapy could ablate the entire volume of cancer. The likelihood of success in this regard may be approximated from the prostate-specific antigen (PSA) failure–free (bNED) survival after RP. Identifying these patients prior to therapy, on the basis of the readily available pretreatment prognostic factors was the goal of this study.
Data from the surgical literature (1, 2) have suggested that the cancer-specific survival after radical prostatectomy (RP) decreases as one advances from pathologic organ– confined to specimen-confined (extracapsular extension [ECE] with negative surgical margins) to positive-margin, seminal vesicle–positive, and finally lymph node–positive disease. These decrements in cancer-specific survival could be due to local failure that in time metastasizes and becomes a distant failure. However, a significant proportion of these deaths are likely due to the increasing incidence of occult distant micrometastatic disease present in men with a more advanced pathologic stage at the time of the RP.
Reprint requests to: Anthony V. D’Amico, M.D., Ph.D., Joint Center for Radiation Therapy, Harvard Medical School, 330 Brookline Avenue–5th floor, Boston, MA 02215. E-mail: [email protected]
Accepted for publication 22 July 1999.
1227
1228
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 45, Number 5, 1999
Table 1. Clinical pretreatment and postoperative pathologic characteristics of the 1742 surgically managed patients studied using the logistic regression multivariable analyses Clinical characteristic
Number
Pathologic OC
Pathologic SC
Pathologic M⫹ or SV⫹
AJCC clinical T1c AJCC clinical T2a AJCC clinical T2b AJCC clinical T2c Biopsy Gleason 2–4 Biopsy Gleason 5–6 Biopsy Gleason 7 Biopsy Gleason 8–10 PSA ⱕ 4 ng/ml PSA ⬎ 4–10 ng/ml PSA ⬎ 10–20 ng/ml PSA ⬎ 20 ng/ml
688 653 214 187 280 999 343 120 184 1044 367 147
529 (77%) 455 (70%) 118 (55%) 89 (48%) 224 (80%) 727 (73%) 185 (54%) 55 (46%) 151 (82%) 786 (75%) 206 (56%) 48 (33%)
65 (9%) 85 (13%) 30 (14%) 23 (12%) 22 (8%) 109 (11%) 59 (17%) 13 (11%) 15 (8%) 113 (11%) 53 (15%) 22 (15%)
94 (14%) 113 (17%) 66 (31%) 75 (40%) 34 (12%) 163 (16%) 99 (29%) 52 (43%) 18 (10%) 145 (14%) 108 (29%) 77 (52%)
Abbreviations: AJCC ⫽ 1992 American Joint Commission on Cancer Staging System; PSA ⫽ prostate-specific antigen; SC ⫽ specimen-confined; OC ⫽ organ-confined; SV ⫽ seminal vesicle; M ⫽ margin; ⫹ ⫽ Positive.
METHODS AND MATERIALS Patient population Seventeen hundred and forty-two men treated with a RP and bilateral pelvic lymph node dissection at the Hospital of the University of Pennsylvania (HUP) or the Brigham and Women’s (B&W) Hospital between 1990 and 1998 who had PSA detected or clinically palpable prostate cancer comprised the study population. Table 1 lists the preoperative clinical and postoperative pathological characteristics of the 1742 study patients. Preoperative staging In all cases, staging evaluation included a history and physical examination including a digital rectal exam (DRE), serum PSA, computed tomography (CT) of the pelvis or an endorectal and pelvic coil magnetic resonance imaging (MRI) scan of the prostate and pelvis, bone scan, and a transrectal ultrasound guided (TRUS) needle biopsy of the prostate with Gleason score histologic grading (3). The prostate biopsy was performed using an 18-gauge Tru-Cut needle (Travenol Laboratories, Deerfield, IL) via a transrectal approach. The clinical stage was obtained from the DRE findings using the current 1992 AJCC staging system (4). Radiologic and biopsy information were not used to determine clinical stage. The PSA was obtained on an ambulatory basis prior to radiologic studies and biopsy. All PSA measurements (5) were made using the Hybritech, Tosoh, or Abbot assays. Treatment and pathologic processing A referee genitourinary pathologist reviewed the diagnostic biopsy specimens for all patients undergoing surgery at the Hospital of the University of Pennsylvania (JET) or the Brigham and Women’s Hospital (AAR). Surgical treatment consisted of a radical retropubic prostatectomy and bilateral pelvic lymph node sampling. If the intraoperative frozen sections of any sampled lymph node were positive for carcinoma, the radical prostatectomy was aborted. Prosta-
tectomy specimens were weighed, measured, inked over the entire surface, and fixed in 10% buffered formalin. Both the apical and basal margins were amputated to a thickness of 5 mm and sectioned parasagitally in a direction perpendicular to the initial transverse incision at 3-mm intervals. The base of the seminal vesicles and the basal cross section were submitted separately for microscopic analysis. The prostate was sectioned perpendicular to the long axis (apical– basal) of the gland along the posterior/rectal surface at 5-mm intervals, with most specimens requiring 4 to 7 cross sections to be entirely sectioned. For each cross section, a single section each of the right and left posterior region was submitted. Finally, a single section from the mid anterior prostate was submitted for microscopic evaluation. Evidence of extraprostatic disease including seminal vesicle invasion (SVI), ECE, and/or positive surgical margins (PSM) were recorded. Disease extending into but not through the prostatic capsule was considered negative for ECE. Follow-up The median follow-up of the 960 and 782 surgically managed patients at HUP or the B&W Hospital was 46 months (8 –100 months) and 42 months (12–96 months) respectively. The patients were seen 1 month postoperatively, then at 3-month intervals for 2 years, every 6 months for 5 years, and annually thereafter. At each follow-up a serum PSA was obtained prior to performing the DRE. All pretreatment PSA values were obtained within 1 month of the date of surgery. No patient was lost to follow-up. Statistical analyses A logistic regression multivariable analysis (6) evaluating the ability of the preoperative PSA, biopsy Gleason score, and the 1992 AJCC clinical T-stage to predict pathologic organ– confined, specimen-confined (ECE with a negative surgical margin), and more advanced disease (margin-positive and/or seminal vesicle–positive) was performed. The
Prostate cancer patient selection for dose escalation
PSA, biopsy Gleason score, and 1992 AJCC stages were first treated as continuous variables and then in a separate analysis treated as categorical variables. In this latter analysis, PSA was grouped as ⬎0 – 4, ⬎4 –10, ⬎10 –15, ⬎15– 20, and ⬎20 ng/ml. Biopsy Gleason score was grouped as 2– 4, 5– 6, 7, and 8 –10. Finally the 1992 AJCC clinical T-stage was grouped as T1c, T2a, T2b, and T2c. Baseline groups for the categorical variables to which other groups were compared were PSA ⬎ 0 – 4 ng/ml, biopsy Gleason 2– 4, and 1992 AJCC clinical stage T1c, 2a. PSA failure was scored when two consecutive detectable PSA values were obtained postoperatively after an undetectable value. The time of PSA failure was then recorded at the time of the first detectable PSA. Time zero was defined as the day of surgery. If a PSA never became undetectable postoperatively, then PSA failure was defined to be at time equal to zero. For illustrative purposes, PSA failure–free survival (bNED) as a function of the pretreatment PSA, biopsy Gleason score, and 1992 AJCC clinical T-stage was estimated using an actuarial calculation according to the method of Kaplan and Meier (7) and graphically displayed. Patients found to have positive pelvic lymph nodes at the time of frozen section or whose pelvic lymph nodes were negative at the time of frozen section but positive on the final sections were started on androgen suppression therapy. Therefore, while these men (n ⫽ 12) were excluded from the Kaplan-Meier analysis of PSA survival, they were included in the margin-positive or SV category for the purpose of the logistic regression analysis.
RESULTS Clinical predictors of pathologic stage The significant negative predictors of pathologic organ– confined disease were analogous to the positive predictors of positive margins or seminal vesicle invasion. These clinical predictors included a PSA ⬎ 10 ng/ml ( p ⬍ 0.0001), biopsy Gleason score ⱖ 7 ( p ⱕ 0.0004), and ⱖ T2b disease ( p ⱕ 0.03). Only biopsy Gleason score 7 ( p ⫽ 0.0006) and PSA ⬎ 10 –15 ng/ml ( p ⫽ 0.04) were significant predictors of specimen-confined disease. Table 2 lists the complete set of p-values for the logistic regression analyses that evaluated the ability of the PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage to predict the various pathologic endpoints. Listed in Table 2 are the results when the clinical indicators were examined as either continuous or categorical variables. Table 3 rank orders the combinations of PSA, biopsy Gleason score, and 1992 AJCC clinical T-stage beginning with those combinations in Groups 1–5 that represent the patients who are most likely to have pathologic organ– confined disease and remain PSA failure–free (bNED) at 5 years after RP. These patients are followed by patients in groups 6 –9 who were more likely to have margin-positive disease or seminal vesicle invasion and least likely to remain PSA failure–free at 5 years following RP.
● A. V. D’AMICO et al.
1229
Table 2. P-values derived from the logistic regression multivariable analysis for each of the pathologic endpoints Clinical characteristic PSA (continuous) Biopsy Gleason score (continuous) 1992 AJCC clinical T-stage (continuous) PSA 0–4 ng/ml PSA ⬎ 4–10 ng/ml PSA ⬎ 10–15 ng/ml PSA ⬎ 15–20 ng/ml PSA ⬎ 20 ng/ml Biopsy Gleason 2–4 Biopsy Gleason 5–6 Biopsy Gleason 7 Biopsy Gleason 8–10 1992 AJCC T1c or T2a 1992 AJCC T2b 1992 AJCC T2c
Organconfined*
Specimen confined
Margin or SV ⫹
⬍ 0.0001
0.004
⬍ 0.0001
⬍ 0.0001
0.006
⬍ 0.0001
⬍ 0.0001 —† 0.11 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001 — 0.06 ⬍ 0.0001 0.0004 — 0.007 0.03
0.33 — 0.20 0.04 0.16 0.26 — 0.22 0.0006 0.84 — 0.31 0.97
0.0001 — 0.27 ⬍ 0.0001 ⬍ 0.0001 ⬍ 0.0001 — 0.28 0.0001 0.0001 — 0.01 0.01
* All significant p-values for this pathologic group had negative coefficients meaning that the likelihood of having pathologic organ– confined disease decreases significantly with an increasing value of the clinical predictor. † Denotes a baseline group for the categorical variable.
bNED Survival stratified by clinical and pathologic factors Figure 1 illustrates the decrement in the estimated 5-year bNED survival with increasing pathologic stage. Specifically, the estimated 5-year bNED rates were 77%, 53%, and 20% for patients with organ-confined, specimen-confined, or at least margin-positive disease. All pairwise p-values for comparisons among these groups of patients were ⬍0.0001. Figure 2 illustrates a similar trend for decreasing estimates of bNED survival as one moves from a group of patients whose clinical pretreatment factors as shown in Table 3 are most often associated with pathologic organ– confined disease (Group 1) to those most often associated with at least margin-positive disease (Groups 6 –9). Specifically, the 5-year bNED rates for patients in Groups 1, 2–5, and 6 –9 are 80%, 62%, and 35% respectively. The pairwise p-values for all comparisons in Fig. 2 were also ⬍0.0001. The justification for the groupings in Figure 2 is shown in Fig. 3 where the estimates of the 5-year bNED survival after RP are 80% for Group 1, 68%, 53%, 68%, and 58% ( p ⬎ 0.05) for Groups 2, 3, 4, and 5 respectively and 37%, 37%, 39%, and 33% ( p ⬎ 0.05) for Groups 6, 7, 8, and 9 respectively. An additional subgroup analysis for each grouping (1 vs. 2–5 vs 6 –9) was performed stratified by pathologic stage to assess PSA outcome in these groups. Figures 4 – 6 depict the bNED survival in these groups for men with organ-confined, specimen-confined, and margin- or SV-positive disease respectively. The 4-year bNED rates for patients in groups 1 vs. 2–5 vs. 6 –9 were 87%, 75%, 70% for men with pathologic organ– confined disease. Similarly these values
1230
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 45, Number 5, 1999
Table 3. Rank ordering of the pretreatment clinical characteristic groups by the frequency of the pathologic outcomes and 5-year PSA failure–free (bNED) survival Group
PSA
b Gleas
T-stage
N
OC
SC
M⫹ or SV⫹*
5-yr bNED
1 2 3 4 5 6 7 8 9
ⱕ10 ⱕ10 ⬎10–15 ⱕ10 ⱕ10 ⬎10–15 ⬎10–15 ⬎10–15
ⱕ6 7 ⱕ6 ⱕ6 7 ⱕ6 7 7 ⬎15 or ⱖ8 or 2c
1c, 2a 1c, 2a 1c, 2a 2b 2b 2b 1c, 2a 2b
801 179 144 87 27 22 34 10 438
81% 67% 70% 76% 52% 39% 35% 40% 48%
9% 20% 13% 7% 26% 23% 24% 20% 14%
10% 13% 17% 17% 22% 38% 41% 40% 38%
80% 68% 53% 68% 58% 37% 37% 39% 33%
* Twelve patients had lymph node ⫹ disease and were included in the M⫹ or SV⫹ category. Abbreviations: b Gleas ⫽ Biopsy Gleason score; T-stage ⫽ 1992 AJCC clinical T-stage; N ⫽ number; OC ⫽ organ-confined; SC ⫽ specimen-confined; M ⫽ margin; SV ⫽ seminal vesicle; ⫹ ⫽ positive.
were 75%, 62%, 37% and 58%, 42%, 15% for men with specimen-confined and margin- or SV-positive disease respectively.
Retrospective comparisons of patients managed with 3D dose escalation conformal external beam RT (8 –10) or external beam ⫹ interstitial implant boost (11, 12) versus conventional-dose external beam RT have reported statistically significant improvements in bNED survival rates. One report from Hanks and colleagues (8) has suggested an improvement in overall survival for men treated with 3D dose escalation RT versus historical controls treated with external beam RT to conventional dose. While these findings await prospective validation, identifying patients today who are most likely to derive a survival benefit from the use of dose escalation techniques is a clinically relevant issue.
Selecting the ideal patient for dose escalation radiation therapy would have two requirements. First, the probability of seminal vesicle invasion, margin-positive, or lymph node–positive disease would need to be low in order to minimize the risk of occult micrometastatic disease. In this study as in others (1, 2) bNED survival after RP was found to be significantly less for patients with specimen-confined, margin-positive disease, or seminal vesicle invasion when compared to pathologic organ– confined and margin-negative disease (Fig. 1). Second, in order to ensure that local only therapy is sufficient for cure, 5-year estimates of bNED survival after RP (Table 3) could be used to approximate success. Using bNED survival after RP as the surrogate for measuring the likelihood of success is reasonable considering that the margins obtained in a RP specimen are less than or equal to the margins used when utilizing higher RT doses either with 3D conformal dose escalation or external beam RT plus an interstitial radiation therapy boost. Specifically,
Fig. 1. PSA (bNED) survival stratified by pathologic organ– confined (OC), specimen-confined (SC), or at margin (M) or seminal vesicle (SV)-positive disease for the 1742 study patients. All pairwise p-values ⬍ 0.0001.
Fig. 2. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3. All pairwise p-values ⬍ 0.0001.
DISCUSSION
Prostate cancer patient selection for dose escalation
● A. V. D’AMICO et al.
1231
Fig. 3. PSA (bNED) survival stratified by specific groupings of the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3.
Fig. 5. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3 for patients with pathologic specimen– confined (SC) disease.
a negative margin in a RP specimen can range from less than 1 mm to as much as 5 mm. This is similar to the smaller RT field margins used at the prostatic–rectal interface (⬃5 mm) when escalating RT dose (9) beyond the conventionaldose range. An additional complexity of smaller margins however, is that the intrinsic prostate motion has been reported to be on the order of 2–3 mm in the anterior– posterior direction (13) which can also impact on the volume of the prostate receiving the prescription dose. Therefore, requiring both that the estimates of 5-year bNED survival after RP are reasonable (⬎50%) and that the patient has a low risk of occult micrometastatic disease could translate into an improvement in survival.
Defining the optimal candidates for dose escalation RT Evaluating the results of Table 3 and Figure 3 provides the basis on which to define 3 distinct patient subgroups. These 3 groups can be defined using the 5-year bNED survival rates and pathologic characteristics found upon pathologic evaluation of the RP specimen. Specifically, patients in Groups 6 –9 who had the lowest 5-year bNED rates (33–39%) also had the highest rates of margin- or SV-positive disease (38 – 41%). These patients would be expected to have the highest risk of occult micrometastatic disease and therefore are least likely to derive a survival benefit from attempts at improving local therapy. Evaluating the impact of systemic therapy and improved local therapies in patient Groups 6 –9 would be reasonable. Such
Fig. 4. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3 for patients with pathologic organ– confined (OC) disease.
Fig. 6. PSA (bNED) survival stratified by subgroups based on the pretreatment PSA level, biopsy Gleason score, and 1992 AJCC clinical T-stage as shown in Table 3 for patients with marginpositive or seminal vesicle–positive (M or SV ⫹) disease.
1232
I. J. Radiation Oncology
●
Biology
●
Physics
an endeavor is currently being performed by the Radiation Therapy Oncology Group in protocol 9408. RTOG 9408 is a Phase III trial evaluating the impact on survival of adding 4 months of total androgen suppression therapy to conventional dose external beam RT. It is interesting to note, however, that of all men in Groups 6 –9, 58% remained PSA failure–free 4 years following RP if they had organ-confined disease as noted in Fig. 4. This cohort of men with organ-confined disease comprised 47% of this high-risk subgroup. Therefore, attempts at improving local therapy through dose escalation in the high-risk patients is likely to improve outcome beyond the improvement noted when androgen suppression therapy is added to conventional-dose external beam radiation therapy (14). The second patient subgroup is comprised of the men in Group 1. These patients had the highest 5-year bNED rate (81%) and the lowest rate of margin- or SV-positive disease (10%). Therefore, these patients would be expected to have the lowest rate of occult micrometastatic disease as well as the smallest local tumor burden. Recently, patients in this subgroup have been suggested by Pinover and colleagues (10) to derive a benefit in bNED survival through the use of 3D dose escalation RT beyond 71.5 Gy. Patients comprising Groups 2–5 represent the final patient subgroup. The 5-year bNED survival of these men was intermediate between the patients in the most favorable (Group 1) and least favorable (Groups 6 –9) groups ranging from 53– 68%. Similarly the finding of margin- or SVpositive disease was also intermediate between the most and least favorable groups, being 13–22%. While some men in this group will have occult micrometastatic disease and will have failed RP on that basis, a significant fraction could
Volume 45, Number 5, 1999
have had a larger volume of local disease which may have been inadequately addressed by the limited margins that can be obtained using RP. In that case, these men would be the optimal candidates for investigating the impact that escalating the RT dose has on survival. In particular, as noted in Fig. 5, 75% and 62% of patients in Groups 2–5 remained PSA failure–free 4 years following RP if they had organconfined (n ⫽ 299) or specimen-confined (n ⫽ 61) disease respectively as noted in Fig. 5. These 360 patients comprised the vast majority (83%) of the whole cohort (n ⫽ 433) providing further evidence that dose escalation in this subgroup may improve outcome in the majority of these patients. Previous retrospective comparisons (15, 16) have reported comparable PSA failure–free survival for the patient groups analyzed in this study who were managed with either a RP or conventional-dose external beam RT. Therefore, it is reasonable to hypothesize that higher RT doses in the select patients whose risk of occult micrometastatic disease is low (Groups 1–5) may improve survival beyond that obtained using RP or conventional-dose RT. CONCLUSION Patients most likely to derive a survival benefit from the improved local control possible using dose escalation techniques in this study were those who had both a low risk of having occult micrometastatic disease (⬍25% M⫹ or SV⫹) and a reasonable likelihood of remaining disease-free after RP (⬎50% 5-year bNED). These patients included those having T1c, 2a, PSA ⬎ 10 –15 ng/ml, and biopsy Gleason ⱕ6 or T1c, 2a, 2b, PSA ⱕ 10 ng/ml, and biopsy Gleason ⱕ7 prostate cancer.
REFERENCES 1. Iselin CE, Robertson JE, Paulson DF. Radical perineal prostatectomy: Oncological outcome during a 20-year period. J Urol 1999;161:163–168. 2. Walsh PC, Partin AW, Epstein JI. Cancer control and quality of life following anatomical radical retropubic prostatectomy: Results at 10 years. J Urol 1994;152:1831–1836. 3. Gleason DF and the Veterans Administration Cooperative Urological Research Group. Histologic grading and staging of prostatic carcinoma. In: Tannenbaum M, editor. Urologic pathology. Philadelphia: Lea & Febiger; 1977. p. 171–187. 4. Beahrs OH, Henson DE, Hutter RVP. American Joint Committee on Cancer, Manual for staging cancer, 4th ed. Philadelphia: J.P. Lippincott; 1992. 5. Oesterling JE, Jacobsen SJ, Klee GG, et al. Free, complexed and total serum prostate specific antigen: The establishment of appropriate reference ranges for their concentrations and ratios. J Urol 1995;154:1090 –1095. 6. Simultaneous inferences and other topics in regression analysis-1. In: Neter J, Wasserman W, Kutner M, editors. Applied linear regression models, 1st ed. Homewood, IL: Richard D. Irwin, Inc; 1983. p. 150 –153. 7. Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Amer Stat Assoc 1958;53:457–500. 8. Hanks GE, Hanlon AL, Horwitz E, Pinover W, Schultheiss T. A relationship of local prostate dose and subsequent metasta-
9.
10.
11.
12. 13.
sis with 3D CRT in prostate cancer (Abstract 38). Int J Radiat Oncol Biol Phys 1998;42S:143. Fiveash JB, Hanks GE, Roach M III, Wang S, Vigneault E, McLaughlin W, Sandler HM. 3D Conformal radiation therapy (3DCRT) for high grade prostate cancer: A multi-institutional review (Abstract 37). Int J Radiat Oncol Biol Phys 1998;42S: 143. Pinover W, Hanlon AL, Horwitz EM, Hanks GE. Defining the appropriate dose for prostate cancer patients with PSA ⬍ 10 ng/ml (Abstract 36). Int J Radiat Oncol Biol Phys 1998;42S: 142. Ragde H, Eigamal AA, Snow PB, Brandt J, Bartolucci AA, Nadir BS, Korb LJ. Ten-year disease free survival after transperineal sonography-guided iodine-125 brachytherapy with or without 45-Gray external beam irradiation in the treatment of patients with clinically localized, low to high Gleason grade prostate carcinoma. Cancer 1998;83:989 –1001. Critz FA, Levinson K, Williams WH, Holladay DA. Prostatespecific antigen nadir: The optimum level after irradiation for prostate cancer. J Clin Oncol 1996;14:2893–2900. Beard CJ, Kijewki P, Bussiere M, Gelman R, Gladstone D, Shaffer K, Plunkett M, Costello P, Coleman CN. Analysis of prostate and seminal vesicle motion: Implications for treatment planning. Int J Radiat Oncol Biol Phys 1996;34:451– 458.
Prostate cancer patient selection for dose escalation
14. Bolla M, Gonsalez D, Warde P, Dubois JB, Mirimanoff R, Storme G, et al. Improved survival in patients with locally advanced prostate cancer treated with radiotherapy and goserelin. N Eng J Med 1997;337:295–300. 15. D’Amico AV, Whittington R, Malkowicz SB, Schultz D, Tomaszewski JE, Kaplan I, Beard C, Wein A. Biochemical outcome after radical prostatectomy, external beam radia-
● A. V. D’AMICO et al.
1233
tion therapy, or interstitial radiation therapy for clinically localized prostate cancer. J Am Med Assoc 1998;280:969 – 974. 16. Keyser D, Kupelian PA, Zippe CD, Levin HS, Klein EA. Stage T1-2 prostate cancer with pre-treatment prostate-specific antigen level ⱕ 10 ng/ml: Radiation therapy or surgery? Int J Radiat Oncol Biol Phys 1997;38:723–729.