Urologic Oncology: Seminars and Original Investigations ] (2014) ∎∎∎–∎∎∎
Original article
PSA response to neoadjuvant androgen deprivation is an independent prognostic marker and may identify patients who benefit from treatment escalation Andrew M. McDonald, M.D., M.S.*, Rojymon Jacob, M.D., Eddy S. Yang, M.D., Ph.D., Michael C. Dobelbower, M.D., Ph.D., Sean Vanlandingham, M.D., John B. Fiveash, M.D. Department of Radiation Oncology, University of Alabama, Birmingham, AL Received 1 August 2013; received in revised form 20 October 2013; accepted 25 October 2013
Abstract Purpose: To determine whether prostate-specific antigen (PSA) measurement after initiation of androgen deprivation therapy (ADT) but prior to the start of radiotherapy (RT) pPSA is an independent predictor of biochemical relapse-free survival (bRFS). We also sought to determine the effect, if any, of factors affecting bRFS for patients who did not achieve pPSA o0.5 ng/mL. Methods and materials: A total of 105 patients with National Comprehensive Cancer Network intermediate- or high-risk prostate cancer treated with neoadjuvant ADT (median ¼ 3.9 mo) and external beam RT had pPSA data available and met the inclusion criteria. Pretreatment and treatment characteristics were included in a Cox proportional hazards model to determine effect on bRFS. Results: Median follow-up was 5.4 years. On multivariable analysis, pPSA Z0.5 ng/mL was associated with worsened bRFS (hazard ratio [HR] ¼ 2.7, P ¼ 0.013). For the subgroup of patients with at most 1 high-risk factor, pPSA remained a statistically significant prognostic factor. For patients within this subgroup who had pPSA Z0.5 ng/mL, the addition of pelvic RT was associated with a trend toward improved outcome (HR ¼ 0.609, P ¼ 0.083). Conclusion: For patients with intermediate- or high-risk prostate cancer receiving neoadjuvant ADT, achieving pPSA o0.5 ng/mL was associated with improved rates of bRFS. Additionally, pPSA measurement may identify patients who may be able to benefit from escalated treatment such as pelvic RT. r 2014 Elsevier Inc. All rights reserved. Keywords: Androgen deprivation; PSA; PSA response to androgen deprivation; External beam radiation; Pelvic radiation
1. Introduction The addition of neoadjuvant androgen deprivation therapy (ADT) to radiotherapy (RT) has been shown to improve rates of biochemical control and biochemical relapse-free survival (bRFS) compared with RT alone for patients with intermediate- or high-risk features [1]. However, neoadjuvant ADT does not appear to improve outcome when combined with radical prostatectomy [2], indicating a unique interaction between ADT and RT. Current theories regarding the mechanism of this interaction include inhibiting the antiapoptotic effects of testosterone and increased T cell-mediated immune response [3]. * Corresponding author. Tel.: þ1-205-934-5670. E-mail address: [email protected] (A.M. McDonald).
1078-1439/$ – see front matter r 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.urolonc.2013.10.019
Given this unique relationship between ADT and RT, the prostate-specific antigen (PSA) response to induction phase of ADT has been suggested as a possible prognostic indicator. Retrospective studies indicate that patients who exhibit a more profound PSA response as a result of ADT experience improved biochemical and clinical outcomes [4–7]. Most recently, McGuire et al. [4] have put forth that the post-ADT, pre-RT PSA (proenzyme PSA [pPSA]) is an independent predictor of bRFS for patients with high-risk disease, with pPSA o0.5 ng/mL being associated with improved bRFS. In this study, we sought to confirm the role of pPSA, particularly pPSA o0.5 ng/mL, as a prognostic indicator in a group of intermediate- and high-risk patients treated with a variety of modern RT techniques, including hypofractionation. We also sought to determine which factors affected
2
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
bRFS for patients not reaching the pPSA o0.5 ng/mL threshold. 2. Methods and materials 2.1. Inclusion criteria The records of every patient receiving external beam RT for clinically localized prostate cancer at UAB since 1999 were reviewed. All patients meeting the following criteria were included in the analysis: biopsy-proven National Comprehensive Cancer Network intermediate- or high-risk prostate cancer (T category of 2b or greater, PSA 410 ng/mL, or Gleason Z7) [8], clinically localized disease, no previous treatment before ADT, definitive external beam RT, and a recorded PSA value at least 30 days from the initiation of ADT but before the start of RT. Data collection was approved by the University of Alabama at Birmingham Institutional Review Board. 2.2. Androgen deprivation All patients received neoadjuvant ADT. ADT was generally begun at least 2 months before RT, with a median of 3.9 months (range 1.3–23.2 mo). The choice of agent was at the discretion of the treating physician; the most common agents were leuprolide, goserelin, or triptorelin depot injections with 4 to 6 weeks of oral bicalutamide or flutamide given initially. In general, patients with high-risk disease were scheduled to receive ADT for 24 months and patients with intermediate-risk disease were scheduled to receive ADT for 6 months. The median total length of ADT across all patients was 24 months (range 3–48 mo). 2.3. Staging The staging process was nearly identical for all patients included in this analysis. For all patients, tissue specimens were reviewed by the UAB Department of Pathology; only Gleason scores reported by the internal review were used for risk-group stratification. The highest pretreatment PSA value was used regardless of where this testing occurred. Digital rectal examination was always performed during the initial consultation; however, imaging modalities such as pelvic magnetic resonance imaging or rectal ultrasound were also used in some cases, at the discretion of the practitioner, to aid in determining the T category. Additionally, all patients were evaluated for metastatic disease by both Tc-99m bone scan and computed tomography of the abdomen and pelvis. 2.4. Radiotherapy A total of 49 patients (47%) received conventionally fractionated (1.8–2.0 Gy per fraction) treatment to a total
prostate dose of 70.2 to 79.0 Gy with a median of 76.3 Gy. Of these 49 patients, 3 were treated as part of the control arm of the RTOG 99-02 trial with structures, fields, and doses as described in the protocol [9]. Of the remaining 46 patients, 38 received pelvic lymph node irradiation which consisted of 45 to 46.8 Gy delivered in 1.8 Gy fractions to the pelvis via a 4-field box technique, with the upper field border set at the superior level of the sacroiliac joints and inferior border 2 cm inferior to the prostate. The remaining 8 patients received conventionally fractionated treatment limited to the prostate and seminal vesicles. A total of 56 patients (53%) received a hypofractionated (2.5–3.0 Gy per fraction) treatment schedule delivered via intensity-modulated RT; 2 patients received 70.2 Gy in 27 fractions, 46 received 70 Gy in 28 fractions, 1 received 67.6 Gy in 26 fractions, and 7 received 60 Gy in 20 fractions. Assuming α/β ¼ 1.5, these hypofractionated regimens converted to 77.1 to 82.2 Gy in 2.0 Gy fraction equivalents, and assuming α/β ¼ 10.0 these regimens converted to 65.0 to 73.8 Gy. Simultaneous conventionally fractionated lymph node irradiation was delivered to 35 patients receiving hypofractionated treatment to the prostate. For these patients, the nodal clinical target volume was generated by contouring a 7-mm extension around the internal iliac, external iliac, and common iliac vessels to the level of mid-S1 to approximate a beam aperture at the level of the L5-S1 junction. The nodal planning target volume was then generated by extending the nodal clinical target volume 7 mm in the lateral directions and 9 mm anteriorly and posteriorly. The dose to the nodal volumes was 50.4 Gy delivered in 1.8 Gy fractions. 2.5. Follow-up, end point definition, and statistical methods Following completion of RT, return visits with PSA measurement were scheduled every 3 to 4 months for the first 2 years and every 6 months thereafter. Biochemical failure was defined using the RTOG-ASTRO Phoenix definition as a rise of 2.0 mg/mL above the nadir [10]. The date of failure was recorded as the date of the defining PSA. Statistical analysis of the data was performed using IBM SPSS Statistics 21 software. Concerning pretreatment patient and treatment characteristics, frequencies were compared by the Pearson Chi-square method and means were compared by the independent samples t test or the independent samples Kruskal-Wallis test. Data collection was approved by the University of Alabama at Birmingham Institutional Review Board. bRFS was defined as the interval between initiation of ADT and most recent PSA follow-up or an event, defined as either a biochemical failure or death by any cause. For the calculation of biochemical freedom-from-failure (bFFF), death was not considered an event, and for the calculation of overall survival (OS), biochemical failure was not considered an event. The actuarial rates of bRFS were calculated using the
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
Kaplan-Meier method. Univariate testing for each variable of interest was performed using the Cox proportional hazards modeling, and factors reaching the threshold of P o 0.2 were included in a multivariable model. A subgroup analysis was performed for patients with either intermediate-risk disease or a single high-risk characteristic. From a statistical perspective, this restriction was made to reduce the bias between escalated therapy and more aggressive disease. From a clinical perspective, this subgroup was identified as the patients whose treatment would most likely be altered based on pPSA measurement, as patients with multiple high-risk characteristics would receive maximal therapy regardless of pPSA levels. Treatment to the pelvic lymph nodes, total length of ADT, total prostate dose, and total equivalent prostate dose were identified as treatment factors that could be modified by the treating physician at the time of pPSA measurement. Again, univariate testing for each variable was performed, and factors reaching P o 0.2 were included in a multivariable model.
3. Results 3.1. Pretreatment and treatment characteristics A total of 741 charts were reviewed with 105 patients meeting the inclusion criteria. The median age was 66.8
3
years at the start of RT. Median follow-up was 5.4 years, calculated from the initiation of ADT. All patients were found to be free of metastatic disease; 71 (68%) were classified as high risk and 34 (32%) as intermediate risk. The median length of neoadjuvant ADT was 3.9 months and the median total length of ADT was 24 months. All but 4 patients received at least 2 months of neoadjuvant ADT. An initial course of an oral antiandrogen was included as part of therapy in 83 (79%) patients. The median length from the initiation of ADT to pPSA measurement was 91 (range 30–613) days with a median of 28 (range 0–436) days from pPSA measurement to the start of RT. 3.2. Factors affecting pPSA A list of pretreatment and treatment factors stratified by whether patients reached the pPSA o0.5 ng/mL threshold are listed in Table 1. A total of 49 (47%) reached pPSA o0.5 ng/mL. Of the factors analyzed, the initial PSA (P ¼ 0.019), the use of an oral antiandrogen (P ¼ 0.04), and having more than 60 days of ADT before pPSA measurement (P ¼ 0.011) were found to have a statistically significant difference between the 2 groups. Once reaching 60 days between initiation of ADT and pPSA measurement, additional time was not statistically significant between the 2 groups (P ¼ 0.228). There was no difference in the overall length of neoadjuvant ADT between the 2 groups.
Table 1 Pretreatment and treatment factors stratified by pPSA level pPSA o 0.5 (n ¼ 49) Frequencies (%) T Stage:
PSA:
Gleason
RT Dose Regimen: Pelvic Field: Anti-androgen used: 460 days of ADT prior to PSA measurement? Means (Range) Age at RT start Total Months of ADT Months of neoadjuvant ADT Days from initiation of ADT to pPSA measurement ††
T1 T2 T3-4 o 10 ng/mL 10 – 20 ng/mL 4 20 ng/mL r6 ¼ 7 Z8 Conventional Hypofractionated Yes No Yes No Yes No
19(39%) 19 (39%) 11 (22%) 29 (59%) 8 (16%) 12 (24%) 13 (27%) 17 (35%) 19 (39%) 27 (55%) 22 (45%) 34 (69%) 15 (31%) 43 (88%) 6 (12%) 44(90%) 5 (10%) 68.3 16.3 4.46 98.3
(47 – 83) (3 – 36) (1.6 – 11.5) (35 – 313)
pPSA Z 0.5 (n ¼ 56) 25 (45%) 24 (43%) 7(13%) 18 (32%) 13 (23%) 25 (45%) 13 (23%) 25 (45%) 18 (32%) 22 (39%) 34 (61%) 41 (73%) 15 (27%) 40 (71%) 16 (29%) 39 (58%) 17 (42%) 65.2 18.2 5.81 104.5
(46 – 83) (3 – 48) (1.3 – 23.2) (30 – 613)
Eighty-four patients received 460 days of ADT prior to pPSA measurement. Bolded text implies statistically significant. chi-square. ** Independent samples T-test. † Independent samples Kruskall-Wallis test. * Pearson
p
0.585*
0.019*
0.536*
0.105* 0.665* 0.04* 0.011*
0.055** 0.387** 0.437† 0.409†
4
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
Fig. Kaplan-Meier estimate of (A) biochemical freedom-from-failure, (B) biochemical relapse-free survival, and (C) overall survival stratified by pPSA level.
3.3. Factors affecting outcome end points The Kaplan-Meier estimates of bRFS, bFFF, and OS are presented in Fig. At 5 years, the rate of bRFS was 90.9% for patients with pPSA o0.5 ng/mL compared with 68.8% for patients with pPSA Z0.5 ng/mL (P ¼ 0.015). The difference between the 2 groups was also significant for bFFF (P ¼ 0.04) but did not reach significance for OS (P ¼ 0.164).
A list of pretreatment and treatment factors analyzed for effect on bRFS are presented in Table 2. The only factor reaching the threshold for statistical significance in univariate analysis was pPSA, with pPSA Z0.5 ng/mL associated with HR ¼ 2.7 (P ¼ 0.013). Factors meeting the inclusion threshold for multivariable analysis were pPSA, the total prostate dose, and the total prostate dose with hypofractionated regimens converted to 2.0 Gy
Table 2 The effect of pretreatment and treatment factors on biochemical relapse-free survival N pPSA PSA
PSA (continuous) Gleason
T-stage
No. High-Risk Characteristics Length of ADT Anti-androgen used Hypofractionated RT Length of neoadjuvant ADT (continuos) Prostate dose (continuous) Converted prostate dose assuming α/β ¼ 1.5* (continuous) Whole-Pelvic RT Age
r 0.5 ng/mL 4 0.5 ng/mL o 10 ng/mL 10-20 ng/mL 4 20 ng/mL r6 ¼ 7 Z8 ¼ 1 ¼ 2 Z3 r1 Z2 o 6 months Z 6 months No Yes No Yes
49 (47%) 56 (53%) 47 (45%) 21 (20%) 37 (35%) 105 26 (25%) 42 (40%) 37 (35%) 44 (42%) 43 (41%) 18 (17%) 86 (82%) 19 (18%) 40 (38%) 65 (62%) 22 (21%) 83 (79%) 49 (47%) 56 (53%) 105 105 105
No Yes r 65 4 65
30 (29%) 75 (71%) 42(40%) 63 (60%)
Univariate HR
p
Multivariate HR
p
2.70 [1.18 – 6.17]
0.013
2.482 [1.082 – 5.692]
0.032
1.44 [0.63 – 3.28] 1.40 [0.50 – 3.94] 0.99 [0.98 – 1.01]
0.390 0.522 0.377
0.91 [0.36 – 2.31] 0.80 [0.32 – 1.99]
0.843 0.628
0.991 [0.128 – 7.665] 0.918 [0.106 – 7.927]
0.993 0.918
1.25 [0.83 – 1.89]
0.284
0.625 [0.252 – 1.547]
0.309
0.719 [0.395 – 1.310]
0.281
0.973 [0.650 – 1.457] 1.042 [0.951 – 1.142]
0.894 0.373
0.86 [0.849 – 1.011] 0.894 [0.785 – 1.02]
0.086 0.096
0.922 [0.838 – 1.013] 0.895 [0.799 -1.002]
0.092 0.055
0.821 [0.54 – 1.24]
0.351
1.287 [0.593 – 2.792]
0.524
* Hypofractionated regimens converted to 2 Gy equivalents. Bolded text under univariate analysis implies meets criteria for multivariate analysis and under multivariate implies statistically significant.
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
equivalents assuming α/β ¼ 1.5. In the multivariable model, pPSA Z0.5 ng/mL (HR ¼ 2.482, P ¼ 0.032) was the only factor to meet the threshold for statistical significance, though the converted prostate dose approached significance (HR ¼ 0.895, P ¼ 0.055).
3.4. Subgroup analysis for patients with at most 1 high-risk characteristic For this subgroup of 86 patients, pPSA remained statistically significant and pPSA Z0.5 ng/mL was associated with HR ¼ 2.66 (P ¼ 0.017). The effects of pelvic RT, total length of ADT, Gleason score, prostate dose, and converted prostate dose were analyzed for effect on bRFS. The univariate and multivariable models are presented in Table 3. For patients with pPSA o0.5 ng/mL, no factor was statistically significant on univariate analysis and only the prostate dose met the criteria for inclusion in a multivariable model, therefore this was not performed. For patients with pPSA Z0.5 ng/mL, no factor was significant on univariate analysis. Pelvic RT and Gleason score met the threshold for inclusion in a multivariable model. Neither factor was statistically significant, though pelvic RT approached significance (HR ¼ 0.609, P ¼ 0.083). Additionally, pelvic RT was associated with the use of greater than 6 months of ADT (P ¼ 0.006) and strongly associated with presence of a high-risk characteristic (P o 0.001).
5
4. Discussion Based on the results of multiple prospective trials, neoadjuvant ADT is commonly added to dose-escalated external beam RT for patients with both intermediate- and high-risk prostate cancer. Despite the widespread use of neoadjuvant ADT, no consensus has been reached regarding measurement of the PSA response during the induction phase. PSA response to ADT has recently been reported as a predictor of long-term outcome in a group of high-risk patients treated with dose-escalated external beam RT, with pPSA o0.5 ng/mL suggested as an optimal cutoff [4]. Across all patients included in our analysis, pPSA o0.5 ng/mL was a statistically significant predictor of improved overall bRFS and bFFF. Initial PSA, use of an antiandrogen, and timing of pPSA measurement were found to have a statistically significant effect as to whether pPSA o0.5 ng/ mL was reached before the initiation of RT. Patients whose pPSA was measured before receiving 60 days of ADT were less likely to reach pPSA o0.5; however, additional time did not seem to have an effect. This phenomenon showcases how the PSA kinetics in response to ADT remain poorly understood. The most common gonadotropin releasing hormone analog utilized in this study, leuprolide 22.5 mg depot injection, has been shown to decrease serum testosterone to o50 ng/mL by day 30 of treatment in 95% of patients [11]. Given the proven efficacy of this treatment, a possible explanation for failure to achieve pPSA o0.5 ng/mL is that the traditional castration threshold of
Table 3 The effect of pretreatment and treatment factors on biochemical relapse-free survival for the subgroup of patients with at most 1 high-risk characteristic N pPSA o 0.5 ng/mL (N ¼ 41) Lymph nodes treated Total length of ADT Gleason score
No Yes r 6 months 4 6 months r6 ¼ 7 Z8
13 28 22 19 13 17 11
(32%) (68%) (54%) (46%) (32%) (41%) (27%)
No Yes r 6 months 4 6 months r6 ¼ 7 Z8
13 32 17 28 12 25 8
(29%) (81%) (38%) (62%) (27%) (56%) (18%)
Prostate dose Converted prostate dose assuming α/β ¼ 1.5* pPSA Z 0.5 ng/mL (N ¼ 45) Lymph nodes treated Total length of ADT Gleason score
Prostate dose Converted prostate dose assuming α/β ¼ 1.5* * Hypofractionated
Univariate HR [95% CI]
p
0.985 [0.322 – 3.012]
0.978
0.948 [0.380 – 2.368]
0.909
1.339 4.298 0.881 0.915
[0.119 [0.257 [0.743 [0.575
– – – –
15.074] 71.847] 1.045] 1.457]
0.106
0.695 [0.367 – 1.315]
0.263
[0.583 [0.406 [0.847 [0.753
– – – –
41.235] 27.021] 1.165] 1.072]
p
0.609 [0.347 – 1.070]
0.083
4.586 [0.551 – 38.179] 3.744 [0.445 – 31.490]
0.159 0.224
0.813 0.310 0.147 0.709
0.652 [0.388 – 1.095]
4.902 3.314 0.993 0.898
Multivariate HR [95% CI]
0.143 0.263 0.934 0.234
regimens converted to 2 Gy equivalent dose. Bolded text under univariate analysis implies meets criteria for multivariate analysis.
6
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
testosterone o50 ng/mL may not be adequate for some patients. A recent consensus recommendation has suggested adopting a serum testosterone level of o20 ng/mL as the new castrate level based on comparison with surgical castration with newer assays [12]. Owing to the suggestion that the current ADT standard may not be adequate, routine testosterone measurement during the neoadjuvant phase may be indicated. Additionally, polymorphisms within the androgen receptor have been identified and shown to be correlated with outcome in patients treated with concurrent ADT [13]. These polymorphisms likely also play role in determining the magnitude of PSA response to neoadjuvant ADT. Use of a short course of oral antiandrogen therapy was associated with lower pPSA measurements in our analysis but did not show an effect of bRFS. The reason for this finding is not immediately clear. The role of antiandrogen therapy in clinical practice remains unclear as well. This topic has been investigated by d'Amico et al. [14], who found retrospectively that increasing length of antiandrogen use was associated with improved rates of PSA failure. Given that luteinizing hormone-releasing hormone depot injections suppress testosterone to o50 ng/mL within 30 days, the association of antiandrogen use and improved PSA control supports the notion that a castration level of 50 ng/mL may be too high. However, this study used relatively short courses of ADT at 6 months and was not associated with dose-escalated RT [15]. The appropriate way to tailor neoadjuvant ADT will continue to be an important topic given the association of ADT with metabolic [16] and sexual adverse effects. Another aim of our work was to further explore the clinical utility of pPSA measurement by identifying patients whose treatment could possibly be altered based on this information. Given that only 34 patients with intermediaterisk disease were included, we felt that this was too small a subgroup to analyze separately. As an alternative, we identified patients with at most 1 high-risk characteristic as a large subgroup of interest. In this subgroup, pPSA o0.5 ng/mL remained associated with improved bRFS. No covariate appeared to affect bRFS for patients of this subgroup with pPSA o0.5 ng/mL; however, for those with pPSA Z0.5 ng/mL, the addition of pelvic RT was associated with a trend toward improved bRFS. This trend toward improved bRFS occurred despite a highly significant association between pelvic RT and worse initial prognostic factors. The fact that pelvic RT was also associated with longer courses of adjuvant ADT may have contributed to the trend toward better outcome in this group, although longer courses of ADT did not show an effect when analyzed independently. Despite the interactions between variables, these results suggest that pPSA measurement may identify additional patients who may benefit from escalated treatment. The primary limitations of this study are the retrospective nature of the analysis and the broad inclusion criteria.
The retrospective design of the study likely introduced a bias toward patients with longer periods between initiation of ADT and the start of RT given that collection of pPSA values has not been a standard practice at our institution. However, we feel that the effect of this bias to be minimal given that no association was found between pPSA and longer courses of neoadjuvant ADT so long as they received at least 60 days. The broad inclusion criteria resulted in a heterogeneous group of patients and treatment methods, the effect of which can be noted in the results. For instance, the initial PSA was found not to have an effect on bRFS despite being clearly associated with pPSA level; this was likely owing to a correlation between low initial PSA and the presence of a higher T category or Gleason score, or both, as all patients were of intermediate or high risk. Additionally, the total length of ADT was found not to have an effect on outcome, probably because of longer courses of ADT being associated with higher risk disease. However, the fact that pPSA was strongly associated with outcome despite the variability of pretreatment and treatment characteristics serves to greatly enhance its generalizability across a broad range of clinical practice. Additional limitations include a relatively small sample size, despite more than 700 cases being reviewed, owing to the fact that pPSA information has not been routinely obtained. Serum testosterone measurements were also not available to correlate with pPSA values as this has not been a standard practice at our institution. The total length of ADT was unique among covariates in that it was the only factor that was able to be altered after treatment was initiated. We noted a number of cases where the length of ADT was altered based on patient or clinician preferences but are unsure of the effect, if any, this has on the analysis. Even in the face of these limitations, these data raise a number of questions. Is pPSA simply a measure of the underlying biology or can more aggressive neoadjuvant treatment lead to an improved long-term outcome? What are the kinetics related to testosterone suppression and PSA response and what role do androgen receptor genetics play? Can the adjuvant phase of treatment be reduced in patients with an optimal PSA response? These questions and others should be used to help guide future clinical trials.
5. Conclusion For patients with intermediate or high-risk prostate cancer receiving neoadjuvant ADT, achieving pPSA o0.5 ng/mL is associated with improved rates of bRFS. For patients receiving neoadjuvant ADT, pPSA appears to be a stronger predictor of outcome than initial PSA. Additionally, pPSA measurement may be useful in identifying patients who may be able to benefit from escalated therapy, though this should be confirmed by prospective research.
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
References [1] Pilepich MV, Winter K, Lawton CA, Krisch RE, Wolkov HB, Movsas B, et al. Androgen suppression adjuvant to definitive radiotherapy in prostate carcinoma—long term results of phase III RTOG 85-31. Int J Radiat Oncol Biol Phys 2005;61:1285–90. [2] Schulman CC, Debruyne FM, Forster G, Selvaggi FP, Zlotta AR, Witjes WP. 4-Year follow-up results of a European prospective randomized study on neoadjuvant hormonal therapy prior to radical prostatectomy in T2-3N0M0 prostate cancer. European study group on neoadjuvant treatment of prostate cancer. Eur Urol 2000;38:706–13. [3] Morse MD, McNees DG. Prostate cancer patients treated with androgen therapy develop persistent changes in adaptive immune responses. Hum Immunol 2010;71:496–504. [4] McGuire SE, Lee AK, Cerne JZ, Munsell MF, Levy LB, Kudchadker RJ, et al. PSA response to neoadjuvant androgen deprivation therapy is a strong independent predictor of survival in high-risk prostate cancer in the dose-escalated radiation therapy era. Int J Radiat Oncol Biol Phys 2013;85:e39–46. [5] Ludgate CM, Bishop DC, Pai, Eldridge B, Lim J, Berthelet E, et al. Neoadjuvant hormone therapy and external beam radiation for localized high-risk prostate cancer: the importance of PSA nadir before radiation. Int J Radiat Oncol Biol Phys 2005;62: 1309–15. [6] Mitchell DM, McAleese J, Park RM, Stewart DP, Stranex S, Eakin RL, et al. Failure to achieve PSA level r1 ng/mL after neoadjuvant LHRHa therapy predicts for lower biochemical control rate and overall survival in localized prostate cancer treated with radiotherapy. Int J Radiat Oncol Biol Phys 2007;69:1467–71. [7] Alexander A, Crook J, Jones S, Malone S, Bowen J, Truong P, et al. Is biochemical response more important than duration of neoadjuvant hormone therapy before radiotherapy for clinically localized prostate cancer? An analysis of the 3- versus 8-month randomized trial. Int J Radiat Oncol Biol Phys 2010;76:23–30.
7
[8] National Comprehensive Cancer Network. NCCN clinical practical guidelines in oncology: prostate cancer. Available at: http://www. nccn.org/professionals/physician_gls/pdf/prostate.pdf. [9] Rosenthal SA, Bae K, Pienta KJ, Sobczak ML, Asbell SO, Rajan R, et al. Phase III multi-institutional trial of adjuvant chemotherapy with paclitaxel, estramustine, and oral etoposide combined with long-term androgen suppression and radiotherapy versus long-term androgen suppression plus radiotherapy alone for high-risk prostate cancer: preliminary toxicity analysis of RTOG 99-02. Int J Radiat Oncol Biol Phys 2009;73:672–8. [10] Roach M, Hanks G, Thames H, Schellhammer P, Shipley WU, Sokol GH, et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG-ASTRO Phoenix consensus conference. Int J Radiat Oncol Biol Phys 2006;64:965–74. [11] U.S. Food and Drug Administration. Drugs@FDA: Lupron label and approval history. Approval document. Available at: http://www.accessdata. fda.gov/drugsatfda_docs/nda/pre96/020517_S000_Lupron_Depot.PDF. [12] Djavan B, Eastham J, Gomella L, Tombal B, Taneja S, Dianat SS, et al. Testosterone in prostate cancer: the Bethesda consensus. BJU Int 2012;110:344–52. [13] Abdel-Wahab M, Berkey BA, Krishan A, O’Brien T, Hammond E, Roach M III, et al. Influence of number of CAG repeats on local control in the RTOG 86-10 protocol. Am J Clin Oncol 2006;29:14–20. [14] D'Amico AV, Chen MH, Renshaw A, Loffredo B, Kantoff PW. Risk of prostate cancer recurrence in men treated with radiation alone or in conjunction with combined or less than combined androgen suppression therapy. J Clin Oncol 2008;26:2979–83. [15] D'Amico AV, Manola J, Loffredo M, Rennshaw AA, DellaCroce A, Kantoff PW. Six-month androgen suppression plus radiation therapy vs. radiation therapy alone for patients with clinically localized prostate cancer: a randomized controlled trial. J Am Med Assoc 2004;292:821–7. [16] Saylor PJ, Smith MR. Metabolic complications of androgen deprivation for prostate cancer. J Urol 2013;189(Suppl.1):S34–42.
Original article
PSA response to neoadjuvant androgen deprivation is an independent prognostic marker and may identify patients who benefit from treatment escalation Andrew M. McDonald, M.D., M.S.*, Rojymon Jacob, M.D., Eddy S. Yang, M.D., Ph.D., Michael C. Dobelbower, M.D., Ph.D., Sean Vanlandingham, M.D., John B. Fiveash, M.D. Department of Radiation Oncology, University of Alabama, Birmingham, AL Received 1 August 2013; received in revised form 20 October 2013; accepted 25 October 2013
Abstract Purpose: To determine whether prostate-specific antigen (PSA) measurement after initiation of androgen deprivation therapy (ADT) but prior to the start of radiotherapy (RT) pPSA is an independent predictor of biochemical relapse-free survival (bRFS). We also sought to determine the effect, if any, of factors affecting bRFS for patients who did not achieve pPSA o0.5 ng/mL. Methods and materials: A total of 105 patients with National Comprehensive Cancer Network intermediate- or high-risk prostate cancer treated with neoadjuvant ADT (median ¼ 3.9 mo) and external beam RT had pPSA data available and met the inclusion criteria. Pretreatment and treatment characteristics were included in a Cox proportional hazards model to determine effect on bRFS. Results: Median follow-up was 5.4 years. On multivariable analysis, pPSA Z0.5 ng/mL was associated with worsened bRFS (hazard ratio [HR] ¼ 2.7, P ¼ 0.013). For the subgroup of patients with at most 1 high-risk factor, pPSA remained a statistically significant prognostic factor. For patients within this subgroup who had pPSA Z0.5 ng/mL, the addition of pelvic RT was associated with a trend toward improved outcome (HR ¼ 0.609, P ¼ 0.083). Conclusion: For patients with intermediate- or high-risk prostate cancer receiving neoadjuvant ADT, achieving pPSA o0.5 ng/mL was associated with improved rates of bRFS. Additionally, pPSA measurement may identify patients who may be able to benefit from escalated treatment such as pelvic RT. r 2014 Elsevier Inc. All rights reserved. Keywords: Androgen deprivation; PSA; PSA response to androgen deprivation; External beam radiation; Pelvic radiation
1. Introduction The addition of neoadjuvant androgen deprivation therapy (ADT) to radiotherapy (RT) has been shown to improve rates of biochemical control and biochemical relapse-free survival (bRFS) compared with RT alone for patients with intermediate- or high-risk features [1]. However, neoadjuvant ADT does not appear to improve outcome when combined with radical prostatectomy [2], indicating a unique interaction between ADT and RT. Current theories regarding the mechanism of this interaction include inhibiting the antiapoptotic effects of testosterone and increased T cell-mediated immune response [3]. * Corresponding author. Tel.: þ1-205-934-5670. E-mail address: [email protected] (A.M. McDonald).
1078-1439/$ – see front matter r 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.urolonc.2013.10.019
Given this unique relationship between ADT and RT, the prostate-specific antigen (PSA) response to induction phase of ADT has been suggested as a possible prognostic indicator. Retrospective studies indicate that patients who exhibit a more profound PSA response as a result of ADT experience improved biochemical and clinical outcomes [4–7]. Most recently, McGuire et al. [4] have put forth that the post-ADT, pre-RT PSA (proenzyme PSA [pPSA]) is an independent predictor of bRFS for patients with high-risk disease, with pPSA o0.5 ng/mL being associated with improved bRFS. In this study, we sought to confirm the role of pPSA, particularly pPSA o0.5 ng/mL, as a prognostic indicator in a group of intermediate- and high-risk patients treated with a variety of modern RT techniques, including hypofractionation. We also sought to determine which factors affected
2
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
bRFS for patients not reaching the pPSA o0.5 ng/mL threshold. 2. Methods and materials 2.1. Inclusion criteria The records of every patient receiving external beam RT for clinically localized prostate cancer at UAB since 1999 were reviewed. All patients meeting the following criteria were included in the analysis: biopsy-proven National Comprehensive Cancer Network intermediate- or high-risk prostate cancer (T category of 2b or greater, PSA 410 ng/mL, or Gleason Z7) [8], clinically localized disease, no previous treatment before ADT, definitive external beam RT, and a recorded PSA value at least 30 days from the initiation of ADT but before the start of RT. Data collection was approved by the University of Alabama at Birmingham Institutional Review Board. 2.2. Androgen deprivation All patients received neoadjuvant ADT. ADT was generally begun at least 2 months before RT, with a median of 3.9 months (range 1.3–23.2 mo). The choice of agent was at the discretion of the treating physician; the most common agents were leuprolide, goserelin, or triptorelin depot injections with 4 to 6 weeks of oral bicalutamide or flutamide given initially. In general, patients with high-risk disease were scheduled to receive ADT for 24 months and patients with intermediate-risk disease were scheduled to receive ADT for 6 months. The median total length of ADT across all patients was 24 months (range 3–48 mo). 2.3. Staging The staging process was nearly identical for all patients included in this analysis. For all patients, tissue specimens were reviewed by the UAB Department of Pathology; only Gleason scores reported by the internal review were used for risk-group stratification. The highest pretreatment PSA value was used regardless of where this testing occurred. Digital rectal examination was always performed during the initial consultation; however, imaging modalities such as pelvic magnetic resonance imaging or rectal ultrasound were also used in some cases, at the discretion of the practitioner, to aid in determining the T category. Additionally, all patients were evaluated for metastatic disease by both Tc-99m bone scan and computed tomography of the abdomen and pelvis. 2.4. Radiotherapy A total of 49 patients (47%) received conventionally fractionated (1.8–2.0 Gy per fraction) treatment to a total
prostate dose of 70.2 to 79.0 Gy with a median of 76.3 Gy. Of these 49 patients, 3 were treated as part of the control arm of the RTOG 99-02 trial with structures, fields, and doses as described in the protocol [9]. Of the remaining 46 patients, 38 received pelvic lymph node irradiation which consisted of 45 to 46.8 Gy delivered in 1.8 Gy fractions to the pelvis via a 4-field box technique, with the upper field border set at the superior level of the sacroiliac joints and inferior border 2 cm inferior to the prostate. The remaining 8 patients received conventionally fractionated treatment limited to the prostate and seminal vesicles. A total of 56 patients (53%) received a hypofractionated (2.5–3.0 Gy per fraction) treatment schedule delivered via intensity-modulated RT; 2 patients received 70.2 Gy in 27 fractions, 46 received 70 Gy in 28 fractions, 1 received 67.6 Gy in 26 fractions, and 7 received 60 Gy in 20 fractions. Assuming α/β ¼ 1.5, these hypofractionated regimens converted to 77.1 to 82.2 Gy in 2.0 Gy fraction equivalents, and assuming α/β ¼ 10.0 these regimens converted to 65.0 to 73.8 Gy. Simultaneous conventionally fractionated lymph node irradiation was delivered to 35 patients receiving hypofractionated treatment to the prostate. For these patients, the nodal clinical target volume was generated by contouring a 7-mm extension around the internal iliac, external iliac, and common iliac vessels to the level of mid-S1 to approximate a beam aperture at the level of the L5-S1 junction. The nodal planning target volume was then generated by extending the nodal clinical target volume 7 mm in the lateral directions and 9 mm anteriorly and posteriorly. The dose to the nodal volumes was 50.4 Gy delivered in 1.8 Gy fractions. 2.5. Follow-up, end point definition, and statistical methods Following completion of RT, return visits with PSA measurement were scheduled every 3 to 4 months for the first 2 years and every 6 months thereafter. Biochemical failure was defined using the RTOG-ASTRO Phoenix definition as a rise of 2.0 mg/mL above the nadir [10]. The date of failure was recorded as the date of the defining PSA. Statistical analysis of the data was performed using IBM SPSS Statistics 21 software. Concerning pretreatment patient and treatment characteristics, frequencies were compared by the Pearson Chi-square method and means were compared by the independent samples t test or the independent samples Kruskal-Wallis test. Data collection was approved by the University of Alabama at Birmingham Institutional Review Board. bRFS was defined as the interval between initiation of ADT and most recent PSA follow-up or an event, defined as either a biochemical failure or death by any cause. For the calculation of biochemical freedom-from-failure (bFFF), death was not considered an event, and for the calculation of overall survival (OS), biochemical failure was not considered an event. The actuarial rates of bRFS were calculated using the
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
Kaplan-Meier method. Univariate testing for each variable of interest was performed using the Cox proportional hazards modeling, and factors reaching the threshold of P o 0.2 were included in a multivariable model. A subgroup analysis was performed for patients with either intermediate-risk disease or a single high-risk characteristic. From a statistical perspective, this restriction was made to reduce the bias between escalated therapy and more aggressive disease. From a clinical perspective, this subgroup was identified as the patients whose treatment would most likely be altered based on pPSA measurement, as patients with multiple high-risk characteristics would receive maximal therapy regardless of pPSA levels. Treatment to the pelvic lymph nodes, total length of ADT, total prostate dose, and total equivalent prostate dose were identified as treatment factors that could be modified by the treating physician at the time of pPSA measurement. Again, univariate testing for each variable was performed, and factors reaching P o 0.2 were included in a multivariable model.
3. Results 3.1. Pretreatment and treatment characteristics A total of 741 charts were reviewed with 105 patients meeting the inclusion criteria. The median age was 66.8
3
years at the start of RT. Median follow-up was 5.4 years, calculated from the initiation of ADT. All patients were found to be free of metastatic disease; 71 (68%) were classified as high risk and 34 (32%) as intermediate risk. The median length of neoadjuvant ADT was 3.9 months and the median total length of ADT was 24 months. All but 4 patients received at least 2 months of neoadjuvant ADT. An initial course of an oral antiandrogen was included as part of therapy in 83 (79%) patients. The median length from the initiation of ADT to pPSA measurement was 91 (range 30–613) days with a median of 28 (range 0–436) days from pPSA measurement to the start of RT. 3.2. Factors affecting pPSA A list of pretreatment and treatment factors stratified by whether patients reached the pPSA o0.5 ng/mL threshold are listed in Table 1. A total of 49 (47%) reached pPSA o0.5 ng/mL. Of the factors analyzed, the initial PSA (P ¼ 0.019), the use of an oral antiandrogen (P ¼ 0.04), and having more than 60 days of ADT before pPSA measurement (P ¼ 0.011) were found to have a statistically significant difference between the 2 groups. Once reaching 60 days between initiation of ADT and pPSA measurement, additional time was not statistically significant between the 2 groups (P ¼ 0.228). There was no difference in the overall length of neoadjuvant ADT between the 2 groups.
Table 1 Pretreatment and treatment factors stratified by pPSA level pPSA o 0.5 (n ¼ 49) Frequencies (%) T Stage:
PSA:
Gleason
RT Dose Regimen: Pelvic Field: Anti-androgen used: 460 days of ADT prior to PSA measurement? Means (Range) Age at RT start Total Months of ADT Months of neoadjuvant ADT Days from initiation of ADT to pPSA measurement ††
T1 T2 T3-4 o 10 ng/mL 10 – 20 ng/mL 4 20 ng/mL r6 ¼ 7 Z8 Conventional Hypofractionated Yes No Yes No Yes No
19(39%) 19 (39%) 11 (22%) 29 (59%) 8 (16%) 12 (24%) 13 (27%) 17 (35%) 19 (39%) 27 (55%) 22 (45%) 34 (69%) 15 (31%) 43 (88%) 6 (12%) 44(90%) 5 (10%) 68.3 16.3 4.46 98.3
(47 – 83) (3 – 36) (1.6 – 11.5) (35 – 313)
pPSA Z 0.5 (n ¼ 56) 25 (45%) 24 (43%) 7(13%) 18 (32%) 13 (23%) 25 (45%) 13 (23%) 25 (45%) 18 (32%) 22 (39%) 34 (61%) 41 (73%) 15 (27%) 40 (71%) 16 (29%) 39 (58%) 17 (42%) 65.2 18.2 5.81 104.5
(46 – 83) (3 – 48) (1.3 – 23.2) (30 – 613)
Eighty-four patients received 460 days of ADT prior to pPSA measurement. Bolded text implies statistically significant. chi-square. ** Independent samples T-test. † Independent samples Kruskall-Wallis test. * Pearson
p
0.585*
0.019*
0.536*
0.105* 0.665* 0.04* 0.011*
0.055** 0.387** 0.437† 0.409†
4
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
Fig. Kaplan-Meier estimate of (A) biochemical freedom-from-failure, (B) biochemical relapse-free survival, and (C) overall survival stratified by pPSA level.
3.3. Factors affecting outcome end points The Kaplan-Meier estimates of bRFS, bFFF, and OS are presented in Fig. At 5 years, the rate of bRFS was 90.9% for patients with pPSA o0.5 ng/mL compared with 68.8% for patients with pPSA Z0.5 ng/mL (P ¼ 0.015). The difference between the 2 groups was also significant for bFFF (P ¼ 0.04) but did not reach significance for OS (P ¼ 0.164).
A list of pretreatment and treatment factors analyzed for effect on bRFS are presented in Table 2. The only factor reaching the threshold for statistical significance in univariate analysis was pPSA, with pPSA Z0.5 ng/mL associated with HR ¼ 2.7 (P ¼ 0.013). Factors meeting the inclusion threshold for multivariable analysis were pPSA, the total prostate dose, and the total prostate dose with hypofractionated regimens converted to 2.0 Gy
Table 2 The effect of pretreatment and treatment factors on biochemical relapse-free survival N pPSA PSA
PSA (continuous) Gleason
T-stage
No. High-Risk Characteristics Length of ADT Anti-androgen used Hypofractionated RT Length of neoadjuvant ADT (continuos) Prostate dose (continuous) Converted prostate dose assuming α/β ¼ 1.5* (continuous) Whole-Pelvic RT Age
r 0.5 ng/mL 4 0.5 ng/mL o 10 ng/mL 10-20 ng/mL 4 20 ng/mL r6 ¼ 7 Z8 ¼ 1 ¼ 2 Z3 r1 Z2 o 6 months Z 6 months No Yes No Yes
49 (47%) 56 (53%) 47 (45%) 21 (20%) 37 (35%) 105 26 (25%) 42 (40%) 37 (35%) 44 (42%) 43 (41%) 18 (17%) 86 (82%) 19 (18%) 40 (38%) 65 (62%) 22 (21%) 83 (79%) 49 (47%) 56 (53%) 105 105 105
No Yes r 65 4 65
30 (29%) 75 (71%) 42(40%) 63 (60%)
Univariate HR
p
Multivariate HR
p
2.70 [1.18 – 6.17]
0.013
2.482 [1.082 – 5.692]
0.032
1.44 [0.63 – 3.28] 1.40 [0.50 – 3.94] 0.99 [0.98 – 1.01]
0.390 0.522 0.377
0.91 [0.36 – 2.31] 0.80 [0.32 – 1.99]
0.843 0.628
0.991 [0.128 – 7.665] 0.918 [0.106 – 7.927]
0.993 0.918
1.25 [0.83 – 1.89]
0.284
0.625 [0.252 – 1.547]
0.309
0.719 [0.395 – 1.310]
0.281
0.973 [0.650 – 1.457] 1.042 [0.951 – 1.142]
0.894 0.373
0.86 [0.849 – 1.011] 0.894 [0.785 – 1.02]
0.086 0.096
0.922 [0.838 – 1.013] 0.895 [0.799 -1.002]
0.092 0.055
0.821 [0.54 – 1.24]
0.351
1.287 [0.593 – 2.792]
0.524
* Hypofractionated regimens converted to 2 Gy equivalents. Bolded text under univariate analysis implies meets criteria for multivariate analysis and under multivariate implies statistically significant.
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
equivalents assuming α/β ¼ 1.5. In the multivariable model, pPSA Z0.5 ng/mL (HR ¼ 2.482, P ¼ 0.032) was the only factor to meet the threshold for statistical significance, though the converted prostate dose approached significance (HR ¼ 0.895, P ¼ 0.055).
3.4. Subgroup analysis for patients with at most 1 high-risk characteristic For this subgroup of 86 patients, pPSA remained statistically significant and pPSA Z0.5 ng/mL was associated with HR ¼ 2.66 (P ¼ 0.017). The effects of pelvic RT, total length of ADT, Gleason score, prostate dose, and converted prostate dose were analyzed for effect on bRFS. The univariate and multivariable models are presented in Table 3. For patients with pPSA o0.5 ng/mL, no factor was statistically significant on univariate analysis and only the prostate dose met the criteria for inclusion in a multivariable model, therefore this was not performed. For patients with pPSA Z0.5 ng/mL, no factor was significant on univariate analysis. Pelvic RT and Gleason score met the threshold for inclusion in a multivariable model. Neither factor was statistically significant, though pelvic RT approached significance (HR ¼ 0.609, P ¼ 0.083). Additionally, pelvic RT was associated with the use of greater than 6 months of ADT (P ¼ 0.006) and strongly associated with presence of a high-risk characteristic (P o 0.001).
5
4. Discussion Based on the results of multiple prospective trials, neoadjuvant ADT is commonly added to dose-escalated external beam RT for patients with both intermediate- and high-risk prostate cancer. Despite the widespread use of neoadjuvant ADT, no consensus has been reached regarding measurement of the PSA response during the induction phase. PSA response to ADT has recently been reported as a predictor of long-term outcome in a group of high-risk patients treated with dose-escalated external beam RT, with pPSA o0.5 ng/mL suggested as an optimal cutoff [4]. Across all patients included in our analysis, pPSA o0.5 ng/mL was a statistically significant predictor of improved overall bRFS and bFFF. Initial PSA, use of an antiandrogen, and timing of pPSA measurement were found to have a statistically significant effect as to whether pPSA o0.5 ng/ mL was reached before the initiation of RT. Patients whose pPSA was measured before receiving 60 days of ADT were less likely to reach pPSA o0.5; however, additional time did not seem to have an effect. This phenomenon showcases how the PSA kinetics in response to ADT remain poorly understood. The most common gonadotropin releasing hormone analog utilized in this study, leuprolide 22.5 mg depot injection, has been shown to decrease serum testosterone to o50 ng/mL by day 30 of treatment in 95% of patients [11]. Given the proven efficacy of this treatment, a possible explanation for failure to achieve pPSA o0.5 ng/mL is that the traditional castration threshold of
Table 3 The effect of pretreatment and treatment factors on biochemical relapse-free survival for the subgroup of patients with at most 1 high-risk characteristic N pPSA o 0.5 ng/mL (N ¼ 41) Lymph nodes treated Total length of ADT Gleason score
No Yes r 6 months 4 6 months r6 ¼ 7 Z8
13 28 22 19 13 17 11
(32%) (68%) (54%) (46%) (32%) (41%) (27%)
No Yes r 6 months 4 6 months r6 ¼ 7 Z8
13 32 17 28 12 25 8
(29%) (81%) (38%) (62%) (27%) (56%) (18%)
Prostate dose Converted prostate dose assuming α/β ¼ 1.5* pPSA Z 0.5 ng/mL (N ¼ 45) Lymph nodes treated Total length of ADT Gleason score
Prostate dose Converted prostate dose assuming α/β ¼ 1.5* * Hypofractionated
Univariate HR [95% CI]
p
0.985 [0.322 – 3.012]
0.978
0.948 [0.380 – 2.368]
0.909
1.339 4.298 0.881 0.915
[0.119 [0.257 [0.743 [0.575
– – – –
15.074] 71.847] 1.045] 1.457]
0.106
0.695 [0.367 – 1.315]
0.263
[0.583 [0.406 [0.847 [0.753
– – – –
41.235] 27.021] 1.165] 1.072]
p
0.609 [0.347 – 1.070]
0.083
4.586 [0.551 – 38.179] 3.744 [0.445 – 31.490]
0.159 0.224
0.813 0.310 0.147 0.709
0.652 [0.388 – 1.095]
4.902 3.314 0.993 0.898
Multivariate HR [95% CI]
0.143 0.263 0.934 0.234
regimens converted to 2 Gy equivalent dose. Bolded text under univariate analysis implies meets criteria for multivariate analysis.
6
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
testosterone o50 ng/mL may not be adequate for some patients. A recent consensus recommendation has suggested adopting a serum testosterone level of o20 ng/mL as the new castrate level based on comparison with surgical castration with newer assays [12]. Owing to the suggestion that the current ADT standard may not be adequate, routine testosterone measurement during the neoadjuvant phase may be indicated. Additionally, polymorphisms within the androgen receptor have been identified and shown to be correlated with outcome in patients treated with concurrent ADT [13]. These polymorphisms likely also play role in determining the magnitude of PSA response to neoadjuvant ADT. Use of a short course of oral antiandrogen therapy was associated with lower pPSA measurements in our analysis but did not show an effect of bRFS. The reason for this finding is not immediately clear. The role of antiandrogen therapy in clinical practice remains unclear as well. This topic has been investigated by d'Amico et al. [14], who found retrospectively that increasing length of antiandrogen use was associated with improved rates of PSA failure. Given that luteinizing hormone-releasing hormone depot injections suppress testosterone to o50 ng/mL within 30 days, the association of antiandrogen use and improved PSA control supports the notion that a castration level of 50 ng/mL may be too high. However, this study used relatively short courses of ADT at 6 months and was not associated with dose-escalated RT [15]. The appropriate way to tailor neoadjuvant ADT will continue to be an important topic given the association of ADT with metabolic [16] and sexual adverse effects. Another aim of our work was to further explore the clinical utility of pPSA measurement by identifying patients whose treatment could possibly be altered based on this information. Given that only 34 patients with intermediaterisk disease were included, we felt that this was too small a subgroup to analyze separately. As an alternative, we identified patients with at most 1 high-risk characteristic as a large subgroup of interest. In this subgroup, pPSA o0.5 ng/mL remained associated with improved bRFS. No covariate appeared to affect bRFS for patients of this subgroup with pPSA o0.5 ng/mL; however, for those with pPSA Z0.5 ng/mL, the addition of pelvic RT was associated with a trend toward improved bRFS. This trend toward improved bRFS occurred despite a highly significant association between pelvic RT and worse initial prognostic factors. The fact that pelvic RT was also associated with longer courses of adjuvant ADT may have contributed to the trend toward better outcome in this group, although longer courses of ADT did not show an effect when analyzed independently. Despite the interactions between variables, these results suggest that pPSA measurement may identify additional patients who may benefit from escalated treatment. The primary limitations of this study are the retrospective nature of the analysis and the broad inclusion criteria.
The retrospective design of the study likely introduced a bias toward patients with longer periods between initiation of ADT and the start of RT given that collection of pPSA values has not been a standard practice at our institution. However, we feel that the effect of this bias to be minimal given that no association was found between pPSA and longer courses of neoadjuvant ADT so long as they received at least 60 days. The broad inclusion criteria resulted in a heterogeneous group of patients and treatment methods, the effect of which can be noted in the results. For instance, the initial PSA was found not to have an effect on bRFS despite being clearly associated with pPSA level; this was likely owing to a correlation between low initial PSA and the presence of a higher T category or Gleason score, or both, as all patients were of intermediate or high risk. Additionally, the total length of ADT was found not to have an effect on outcome, probably because of longer courses of ADT being associated with higher risk disease. However, the fact that pPSA was strongly associated with outcome despite the variability of pretreatment and treatment characteristics serves to greatly enhance its generalizability across a broad range of clinical practice. Additional limitations include a relatively small sample size, despite more than 700 cases being reviewed, owing to the fact that pPSA information has not been routinely obtained. Serum testosterone measurements were also not available to correlate with pPSA values as this has not been a standard practice at our institution. The total length of ADT was unique among covariates in that it was the only factor that was able to be altered after treatment was initiated. We noted a number of cases where the length of ADT was altered based on patient or clinician preferences but are unsure of the effect, if any, this has on the analysis. Even in the face of these limitations, these data raise a number of questions. Is pPSA simply a measure of the underlying biology or can more aggressive neoadjuvant treatment lead to an improved long-term outcome? What are the kinetics related to testosterone suppression and PSA response and what role do androgen receptor genetics play? Can the adjuvant phase of treatment be reduced in patients with an optimal PSA response? These questions and others should be used to help guide future clinical trials.
5. Conclusion For patients with intermediate or high-risk prostate cancer receiving neoadjuvant ADT, achieving pPSA o0.5 ng/mL is associated with improved rates of bRFS. For patients receiving neoadjuvant ADT, pPSA appears to be a stronger predictor of outcome than initial PSA. Additionally, pPSA measurement may be useful in identifying patients who may be able to benefit from escalated therapy, though this should be confirmed by prospective research.
A.M. McDonald et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7
References [1] Pilepich MV, Winter K, Lawton CA, Krisch RE, Wolkov HB, Movsas B, et al. Androgen suppression adjuvant to definitive radiotherapy in prostate carcinoma—long term results of phase III RTOG 85-31. Int J Radiat Oncol Biol Phys 2005;61:1285–90. [2] Schulman CC, Debruyne FM, Forster G, Selvaggi FP, Zlotta AR, Witjes WP. 4-Year follow-up results of a European prospective randomized study on neoadjuvant hormonal therapy prior to radical prostatectomy in T2-3N0M0 prostate cancer. European study group on neoadjuvant treatment of prostate cancer. Eur Urol 2000;38:706–13. [3] Morse MD, McNees DG. Prostate cancer patients treated with androgen therapy develop persistent changes in adaptive immune responses. Hum Immunol 2010;71:496–504. [4] McGuire SE, Lee AK, Cerne JZ, Munsell MF, Levy LB, Kudchadker RJ, et al. PSA response to neoadjuvant androgen deprivation therapy is a strong independent predictor of survival in high-risk prostate cancer in the dose-escalated radiation therapy era. Int J Radiat Oncol Biol Phys 2013;85:e39–46. [5] Ludgate CM, Bishop DC, Pai, Eldridge B, Lim J, Berthelet E, et al. Neoadjuvant hormone therapy and external beam radiation for localized high-risk prostate cancer: the importance of PSA nadir before radiation. Int J Radiat Oncol Biol Phys 2005;62: 1309–15. [6] Mitchell DM, McAleese J, Park RM, Stewart DP, Stranex S, Eakin RL, et al. Failure to achieve PSA level r1 ng/mL after neoadjuvant LHRHa therapy predicts for lower biochemical control rate and overall survival in localized prostate cancer treated with radiotherapy. Int J Radiat Oncol Biol Phys 2007;69:1467–71. [7] Alexander A, Crook J, Jones S, Malone S, Bowen J, Truong P, et al. Is biochemical response more important than duration of neoadjuvant hormone therapy before radiotherapy for clinically localized prostate cancer? An analysis of the 3- versus 8-month randomized trial. Int J Radiat Oncol Biol Phys 2010;76:23–30.
7
[8] National Comprehensive Cancer Network. NCCN clinical practical guidelines in oncology: prostate cancer. Available at: http://www. nccn.org/professionals/physician_gls/pdf/prostate.pdf. [9] Rosenthal SA, Bae K, Pienta KJ, Sobczak ML, Asbell SO, Rajan R, et al. Phase III multi-institutional trial of adjuvant chemotherapy with paclitaxel, estramustine, and oral etoposide combined with long-term androgen suppression and radiotherapy versus long-term androgen suppression plus radiotherapy alone for high-risk prostate cancer: preliminary toxicity analysis of RTOG 99-02. Int J Radiat Oncol Biol Phys 2009;73:672–8. [10] Roach M, Hanks G, Thames H, Schellhammer P, Shipley WU, Sokol GH, et al. Defining biochemical failure following radiotherapy with or without hormonal therapy in men with clinically localized prostate cancer: recommendations of the RTOG-ASTRO Phoenix consensus conference. Int J Radiat Oncol Biol Phys 2006;64:965–74. [11] U.S. Food and Drug Administration. Drugs@FDA: Lupron label and approval history. Approval document. Available at: http://www.accessdata. fda.gov/drugsatfda_docs/nda/pre96/020517_S000_Lupron_Depot.PDF. [12] Djavan B, Eastham J, Gomella L, Tombal B, Taneja S, Dianat SS, et al. Testosterone in prostate cancer: the Bethesda consensus. BJU Int 2012;110:344–52. [13] Abdel-Wahab M, Berkey BA, Krishan A, O’Brien T, Hammond E, Roach M III, et al. Influence of number of CAG repeats on local control in the RTOG 86-10 protocol. Am J Clin Oncol 2006;29:14–20. [14] D'Amico AV, Chen MH, Renshaw A, Loffredo B, Kantoff PW. Risk of prostate cancer recurrence in men treated with radiation alone or in conjunction with combined or less than combined androgen suppression therapy. J Clin Oncol 2008;26:2979–83. [15] D'Amico AV, Manola J, Loffredo M, Rennshaw AA, DellaCroce A, Kantoff PW. Six-month androgen suppression plus radiation therapy vs. radiation therapy alone for patients with clinically localized prostate cancer: a randomized controlled trial. J Am Med Assoc 2004;292:821–7. [16] Saylor PJ, Smith MR. Metabolic complications of androgen deprivation for prostate cancer. J Urol 2013;189(Suppl.1):S34–42.
