Advertisement

The dose–response of salvage radiotherapy following radical prostatectomy: A systematic review and meta-analysis

  • Christopher R. King
    Correspondence
    Address: Radiation Oncology & Urology, UCLA School of Medicine, Dept. of Radiation Oncology, 1223 16th Street, Santa Monica, CA 90404, United States.
    Affiliations
    Department of Radiation Oncology, UCLA School of Medicine, Los Angeles, United States
    Search for articles by this author
Published:November 15, 2016DOI:https://doi.org/10.1016/j.radonc.2016.10.026

      Abstract

      Purpose/objectives

      To date neither the optimal radiotherapy dose nor the existence of a dose–response has been established for salvage RT (SRT).

      Materials/methods

      A systematic review from 1996 to 2015 and meta-analysis was performed to identify the pathologic, clinical and treatment factors associated with relapse-free survival (RFS) after SRT (uniformly defined as a PSA > 0.2 ng/mL or rising above post-SRT nadir). A sigmoidal dose–response curve was objectively fitted and a non-parametric statistical test used to determine significance.

      Results

      71 studies (10,034 patients) satisfied the meta-analysis criteria. SRT dose (p = 0.0001), PSA prior to SRT (p = 0.0009), ECE+ (p = 0.039) and SV+ (p = 0.046) had significant associations with RFS. Statistical analyses confirmed the independence of SRT dose–response. Omission of series with ADT did not alter results. Dose–response is well fit by a sigmoidal curve (p = 0.0001) with a TCD50 of 65.8 Gy, with a dose of 70 Gy achieving 58.4% RFS vs. 38.5% for 60 Gy. A 2.0% [95% CI 1.1–3.2] improvement in RFS is achieved for each Gy. The SRT dose–response remarkably parallels that for definitive RT of localized disease.

      Conclusions

      This study provides level 2a evidence for dose-escalated SRT > 70 Gy. The presence of an SRT dose–response for microscopic disease supports the hypothesis that prostate cancer is inherently radio-resistant.

      Keywords

      The eradication of microscopic residual disease, and therefore cure after radical prostatectomy (RP) can potentially be achieved with the use of salvage radiotherapy (SRT). Indeed, an analysis of the patterns of failure from the SWOG adjuvant RT randomized trial demonstrated that treatment failure is predominantly local and that therefore an improvement in local therapy will result in improved outcomes [
      • Thompson I.M.
      • Tangen C.M.
      • Paradelo J.
      • et al.
      Adjuvant radiotherapy for pathologically advanced prostate cancer: A randomized trial.
      ].
      The known prognostic factors revealed by retrospective studies include a positive surgical margin, pathologically organ-confined disease, low Gleason grade, lower PSA level prior to SRT, greater time interval to failure, and longer PSA doubling times [
      • Stephenson A.J.
      • Scardino P.T.
      • Kattan M.W.
      • et al.
      Predicting the outcome of salvage radiation therapy for recurrent prostate cancer after radical prostatectomy.
      ,
      • Stephenson A.J.
      • Shariat S.F.
      • Zelefsky M.J.
      • et al.
      Salvage radiotherapy for recurrent prostate cancer after radical prostatectomy.
      ]. Some of these factors are naturally related to the likelihood of persistent local vs. systemic disease, and some relate to the burden of residual disease or to an aggressive biology.
      Whereas ample evidence exists in support of dose-escalation to 78 Gy or higher (in 1.8–2 Gy per fraction) for definitive radiotherapy of localized prostate cancer [
      • Pollack A.
      • Zagars G.K.
      • Starkschall G.
      • et al.
      Prostate cancer radiation dose-response: results of the M. D. Anderson phase III randomized trial.
      ,
      • Zietman A.L.
      • DeSilvio M.L.
      • Slater J.D.
      • et al.
      Comparison of conventional-dose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate: A randomized controlled trial.
      ,
      • Eade T.N.
      • Hanlon A.L.
      • Horwitz E.M.
      • et al.
      What dose of external beam radiation is high enough for prostate cancer?.
      ,
      • Zelefsky M.J.
      • Sa Leibel
      • Gaudin P.B.
      • et al.
      Dose–escalation with three-dimensional conformal radiation therapy affects the outcome in prostate cancer.
      ,
      • Catton C.
      • Gospodarowicz M.
      • Mui J.
      • et al.
      Clinical and biochemical outcome of conventional dose radiotherapy for localized prostate cancer.
      ] very few studies are even capable of examining the dose–response in the post-operative setting due to the very narrow range in doses that are commonly used. For post-operative radiotherapy the ASTRO-AUA consensus guidelines currently recommend a dose of 64 Gy at conventional dose fractionation [
      • Valicenti R.K.
      • Thompson Jr., I.
      • Albertsen P.
      • et al.
      Adjuvant and salvage radiation therapy after prostatectomy: american society for radiation oncology/american urological association guidelines.
      ]. It is also worth noting that the RCT for adjuvant RT vs. observation prescribed a dose of 60 Gy for EORTC [
      • Bolla M.
      • van Poppel H.
      • Collette L.
      • et al.
      Postoperative radiotherapy after radical prostatectomy: a randomized controlled trial (EORTC trial 22911).
      ] and ARO [
      • Wiegel T.
      • Bottke D.
      • Steiner U.
      • et al.
      Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pT3 prostate cancer with postoperative undetectable prostate-specific antigen: ARO 96–02/AUO AP 09/95.
      ] and 60-64 Gy for SWOG [
      • Thompson I.M.
      • Tangen C.M.
      • Paradelo J.
      • et al.
      Adjuvant radiotherapy for pathologically advanced prostate cancer: A randomized trial.
      ], and that the three ongoing RCT for early salvage vs. adjuvant RT trials: RAVES [

      Trans Tasman Radiation Oncology Group (TROG 08–03) RAVES trial: Radiotherapy – Adjuvant Versus Early Salvage.

      ], RADICALS [
      • Parker C.
      • Clarke N.
      • Logue J.
      • et al.
      RADICALS (Radiotherapy and Androgen Deprivation in Combination after Local Surgery).
      ] and GETUG-17 [
      • Richaud P.
      • Sargos P.
      • Henriques de Figueiredo B.
      • et al.
      Postoperative radiotherapy of prostate cancer (GETUG-17).
      ] prescribe a dose of 64–66 Gy. It is commonly presumed, albeit unproven, that tumor burden in the post-operative setting is microscopic and therefore that lower effective doses are necessary when compared to what is needed in the definitive setting.
      To date, neither the optimal salvage radiotherapy dose nor the existence of a dose–response relationship has been established by randomized clinical trials (save for the SAKK 09/10 trial comparing 64 Gy vs. 70 Gy whose results are maturing [
      • Ghadjar P.
      • Hayoz S.
      • Bernhard J.
      • et al.
      Acute toxicity and quality of life after dose-intensified salvage radiation therapy for biochemically recurrent prostate cancer after prostatectomy: first results of the randomized trial SAKK 09/10.
      ]). Using the methodology of systematic review and meta-analysis, the present study tests the hypothesis that salvage RT dose should be escalated in order to improve the biochemical disease-free survival for these patients.

      Methods

      Study selection criteria

      Following the principles and 27-point checklist set forth by PRISMA – preferred reporting items for systematic reviews and meta-analyses [
      • Moher D.
      • Liberati A.
      • Tetzlaff J.
      • Altman D.G.
      • PRISMA Group
      Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement.
      ], a comprehensive and systematic review of all indexed published reports in the English language of salvage radiotherapy after radical prostatectomy was performed from PubMed. A manual search of all references contained within these publications was also performed to identify any additional studies not identified in that database. To satisfy the uniformity criteria for systematic review, studies were required to: (1) report Kaplan–Meier relapse-free survival (RFS) at minimum 3–5 years; (2) report the median pre-SRT PSA level, the median RT dose, usage of androgen deprivation therapy (ADT); and (3) report all surgical pathologic factors (margin status, extracapsular extension (ECE), seminal vesicle (SV) invasion, pathologic Gleason grade, nodal status). Studies not providing complete information for each factor or reporting only mean values (as distinct from median) were excluded to avoid skewness and outlier effects. Relapse was uniformly defined as a PSA of 0.2 ng/mL or rising above its nadir after SRT. Time to failure was uniformly defined as time from SRT until relapse. All doses were given at conventional fractionation (1.8–2 Gy per daily fraction).

      Statistical analysis

      Clinical, pathologic and treatment factors associated with RFS were analyzed with the Spearman rank correlation test and considered significant when p-value <0.05. The Spearman statistic is ideally suited because as a non-parametric test it does not assume a linear relationship and therefore makes the least assumptions on the data. It yields the correlation coefficient rho (which ranges from −1 to +1) and a p-value. Robustness was performed by using Grubbs’s test to remove outliers at a level of p < 0.05 (two-tailed). Analysis was performed with StatView (v. 5.0.1.0, SAS, Cary, North Carolina).

      Sigmoidal dose–response relationship

      Tumor control vs. dose commonly reveals a sigmoidal relationship, often called a logistic curve. To interpret the observed relationship between dose and RFS, the standard tumor control probability (TCP) equation is used, as given by:
      TCP=e(d-TCD50)/k/[1+e(d-TCD50)/k]


      where d is total dose, TCD50 is the dose to achieve 50% tumor control, and k is a fitting parameter related to the slope at the TCD50 dose point [
      • Okunieff P.
      • Morgan D.
      • Niemierko A.
      • Suit H.D.
      Radiation dose-response of human tumors.
      ]. The proportional gain in TCP per additional Gy within the mid part of the TCP curve is given by the parameter Slope50 = 25/k (in units of % per Gy). PSA relapse-free survival (bNED) is used as the surrogate for tumor control. This dose–response relationship is fitted to the data by the weighted least-squares method using the Solver function within Excel (Microsoft Office 2010).

      Results

      From a review of all studies published between 1996 and 2015 a total of 71 series encompassing 10,034 patients satisfied the entry criteria for systematic review and are itemized in supplementary tables and presented as summary statistics for each factor in Table 1. These series add up to 76 separate table entries as a few series report on several separate subgroups, eg. prostate bed vs. pelvic RT or dose group. Kaplan–Meier RFS was reported at 5 years for 55 series (77%) and at 3 years for the remaining 16 series (23%). Median follow-up had an average of 51.6 ± 21 months, median pre-SRT PSA had an average of 0.84 ± 0.63 ng/mL (range 0.2 to 3.7 ng/mL), median dose had an average of 65.8 ± 3.2 Gy (range 60 to 76 Gy), and RFS was an average of 51 ± 16% (range 14–88%). In terms of short-course concurrent ADT, there were 18 entries (24%) reporting more than a 15% ADT usage rate, 11 entries (14%) with less than 15% ADT usage rate, whereas 47 entries (62%) had no ADT usage at all.
      Table 1Summary of clinical, pathologic and treatment factors for salvage RT series that meet the criteria for systematic review and meta-analysis (71 series, 10,034 patients).
      FUprePSAbNEDTimem+ECE+Dose
      All doses were given with conventional fractionation (1.8–2Gy per fraction).
      ADTGS8-10SV+LN+
      Monthsng/mL%Year%%Gy%%%%
      Mean51.60.8451.04.3555.544.965.810.524.721.40.9
      Min210.2014.02129600000
      Max1673.788.0682.896.67689.77172.49.5
      Std. Dev.20.80.6315.80.914.718.73.218.713.010.52.2
      FU = median follow-up in months, prePSA = median pre-salvage RT PSA, bNED = % biochemical no evidence of disease by Kaplan–Meier, Time = year at which bNED reported, m+ = % positive margins, ECE+ = % + ECE (pT3a), Dose = median salvage RT dose in Gy, ADT = % usage of androgen deprivation therapy, GS8-10 = % with path Gleason 8–10, SV+ = % with positive seminal vesicles (pT3b), LN+ = % with involved nodes, n/a = not available.
      low asterisk All doses were given with conventional fractionation (1.8–2 Gy per fraction).
      Of the treatment factors only SRT dose (p = 0.0001) and PSA level prior to SRT (p = 0.0009), and of the pathologic factors only ECE+ (p = 0.039) and SV+ (p = 0.046) had a significant association with RFS. Within the interval of PSA between 0 and 1, there is a 2.4% decline in PSA RFS for each 0.1 ng/mL increase in PSA.
      Robustness was performed by identifying and eliminating outliers from the analysis using Grubbs’s test. Six outlier series were found (the series corresponding to FU of 167 months, PSA level prior to SRT of 3.7, ADT usage of 90%, GS8-10 of 71%, SV + of 72%, LN + of 10%). After removal of these 6 outliers, SRT dose (p = 0.0006) and PSA level prior to RT (p = 0.0057) remained significant. Because of the known biologic interaction between ADT use and RT, as well as ADT potentially confounding time to PSA relapse, the analysis was repeated after removal of 18 entries where ADT was used among > 15% of patients. After removal of these entries, 56 series remained analyzable [with 8181 patients], with SRT Dose (p = 0.0008) and PSA level prior to RT (0.002) remaining significant. When completely excluding all series with any use of ADT, we are left with 47 analyzable series [with 5374 patients], yielding SRT Dose (p = 0.011) and PSA level prior to RT (p = 0.011) as still significant. These results are summarized in Table 2.
      Table 2Clinical, pathologic and treatment factors associated with relapse-free survival after salvage RT.
      FactorAll StudiesStudies minus outliersStudies minimal ADT
      Studies where more than 15% of patients received ADT were removed.
      Studies w/o ADT
      71 series [10,034 pts]66 series [9672 pts]56 series [8181 pts]47 series [5374 pts]
      rho
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      p-Value
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      rho
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      p-Value
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      rho
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      p-Value
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      rho
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      p-Value
      Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      RT Dose0.4440.00010.4110.00060.4440.00080.3790.011
      Pre-Salvage RT PSA−0.3840.0009−0.3300.0057−0.4060.002−0.3740.011
      m+0.2100.0810.1660.180.1650.220.1900.22
      ECE+−0.2450.039−0.2400.049−0.3080.022−0.2730.073
      ADT−0.0960.41−0.1150.33−0.0360.78nana
      pGS 8–10−0.1640.17−0.2160.084−0.2440.078−0.1890.23
      SV+−0.2420.046−0.2940.019−0.3850.005−0.3290.033
      LN+0.2040.0770.1620.170.2050.120.3340.023
      Bold is used to highlight p-values that are significant (i.e. <0.05).
      +m = positive margins, +ECE = positive extracapsular extension (pT3a), RT Dose = salvage RT dose, ADT = usage of androgen deprivation therapy, pGS8-10 = pathologic Gleason 8–10, SV+ = positive seminal vesicles (pT3b), LN+ = involved nodes.
      a Rho (ranging from −1 to +1) and associated p-value are from the non-parametric Spearman rank correlation test.
      b Studies where more than 15% of patients received ADT were removed.
      Since a multivariate analysis is not mathematically feasible (because individual patient level data are not available) pair-wise correlations between SRT dose and each factor was made to evaluate for any potential interactions. There were no statistically significant correlations found (p > 0.06 to 0.81). There was also no correlation between SRT dose and length of FU (p = 0.89). The lack of statistical interactions between factors is consistent with SRT dose being independently associated with RFS.
      In Fig. 1 is presented the observed dose–response relationship between SRT dose vs. RFS. It is very well fit by the sigmoidal relationship (rho = 0.444, p = 0.0001) with TCD50 = 65.8 Gy, and Slope50 = 2.0% per Gy [95%CI1.1–3.2]. Dose for salvage RT in the range of 60 to 70 Gy appears to be on the middle steep part of the sigmoidal dose–response curve, with a dose of 70 Gy achieving 58.4% RFS vs. 38.5% for 60 Gy.
      Figure thumbnail gr1
      Fig. 1PSA relapse-free survival vs. SRT dose for all eligible salvage RT studies. In this bubble plot the number of patients in each series is reflected by the size of each circle. The data are well fit by a sigmoidal dose–response curve (dashed curve, Spearman’s rho = 0.444, p = 0.0001). The fitted TCD50 = 65.8 Gy, and there’s a 2.0% [95% CI 1.1–3.2] improvement in RFS achieved for each additional Gy. Dose is assumed delivered at conventional fractionation (1.8–2 Gy per daily fraction).

      Discussion

      This study identifies two of the most influential factors within the context of SRT: namely RT dose and PSA level prior to SRT. Two previous studies [
      • King C.R.
      The timing of salvage radiotherapy after radical prostatectomy: a systematic review.
      ,
      • Ohri N.
      • Dicker A.P.
      • Trabulsi E.J.
      • Showalter T.N.
      Can early implementation of salvage radiotherapy for prostate cancer improve the therapeutic ratio? A systematic review and regression meta-analysis with radiobiological modelling.
      ] had focused primarily on the timing of SRT, although they had strongly suggested a dose–response was present. The updated ‘Stephenson’ salvage RT nomogram [
      • Tendulkar R.D.
      • Agarwal S.
      • Gao T.
      • et al.
      Contemporary update of a multi-institutional predictive nomogram for salvage radiotherapy after radical prostatectomy.
      ] shows a biochemical relapse-free survival advantage on multivariate analysis comparing SRT dose of ⩾66 Gy vs <66 Gy. While our current study also confirms the findings of the aforementioned studies of the advantage of delivering SRT at the lowest possible PSA and using higher doses, our current study’s emphasis is on analyzing RT dose–response. While our study does not imply that the other prognostic and treatment factors (pathologic stage, margin status, Gleason grade, or use of ADT) are not relevant, but rather that they are dominated by these other two factors, PSA and dose. Our main finding is the existence of a salvage RT dose–response, independent of other factors, and that dose–response is well fit by a sigmoidal curve (p = 0.0001) with a TCD50 of 65.8 Gy. A 2.0% [95% CI 1.1–3.2] improvement in RFS is achieved for each additional Gy within the dose range of 60 Gy to 70 Gy. This dose range appears to lie within the steep part of the dose–response curve.
      This large systematic review and meta-analysis of retrospective studies provides level 2a evidence, as defined by the Oxford Centre for Evidence Based Medicine [

      Howick J, Chalmers I, Glasziou P, et al. Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence (OCEBM LoE Working Group) (http://www.cebm.net/wp-content/uploads/2014/06/CEBM-Levels-of-Evidence-2.1.pdf).

      ] for escalating salvage RT dose to at least 70 Gy and very likely higher.
      It is commonly presumed, albeit unproven, that because tumor burden in the post-operative setting is microscopic, that lower effective doses are necessary when compared to what is needed in the definitive setting. So in order to appreciate the relevance of a dose–response for SRT, a similar analysis was performed for definitive RT alone (no ADT) of localized prostate cancer. The definitive RT series used include the University of Texas M.D. Anderson (MDA) dose-escalation trial [
      • Pollack A.
      • Zagars G.K.
      • Starkschall G.
      • et al.
      Prostate cancer radiation dose-response: results of the M. D. Anderson phase III randomized trial.
      ], the Massachusetts General Hospital (MGH) dose-escalation trial [
      • Zietman A.L.
      • DeSilvio M.L.
      • Slater J.D.
      • et al.
      Comparison of conventional-dose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate: A randomized controlled trial.
      ], the prospective dose-escalation series from Fox Chase Cancer Center (FC) [
      • Eade T.N.
      • Hanlon A.L.
      • Horwitz E.M.
      • et al.
      What dose of external beam radiation is high enough for prostate cancer?.
      ] and Memorial Sloan Kettering Cancer Center (MSK) [
      • Zelefsky M.J.
      • Sa Leibel
      • Gaudin P.B.
      • et al.
      Dose–escalation with three-dimensional conformal radiation therapy affects the outcome in prostate cancer.
      ], and the series from Princess Margaret Hospital (PMH) [
      • Catton C.
      • Gospodarowicz M.
      • Mui J.
      • et al.
      Clinical and biochemical outcome of conventional dose radiotherapy for localized prostate cancer.
      ]. The fitted dose–response yields a TCD50 of 65.9 Gy with a 2.6% per Gy slope. The curve for definitive RT is shown in Fig. 2, with the SRT dose–response curve from the current study superimposed. That the SRT dose–response so remarkably parallels that for definitive RT of localized disease supports the hypothesis that prostate cancer is inherently radio-resistant.
      Figure thumbnail gr2
      Fig. 2PSA relapse-free survival vs. RT dose for definitive RT studies. A sigmoidal dose–response curve (solid curve) is fit to the data, yielding a TCD50 = 65.9 Gy and a 2.6% improvement in RFS achieved for each additional Gy. The dose–response curve for salvage RT (dashed curve) from is superimposed for comparison. Dose is assumed delivered at conventional fractionation (1.8–2 Gy per daily fraction).
      The findings from this study suggest that the recommended and trial salvage doses were all considerably too low, being barely at the TCD50. The aforementioned include the ASTRO-AUA consensus guidelines [
      • Valicenti R.K.
      • Thompson Jr., I.
      • Albertsen P.
      • et al.
      Adjuvant and salvage radiation therapy after prostatectomy: american society for radiation oncology/american urological association guidelines.
      ] – 64 Gy, or that used in the RCT for adjuvant RT vs. observation – 60 Gy for EORTC [
      • Bolla M.
      • van Poppel H.
      • Collette L.
      • et al.
      Postoperative radiotherapy after radical prostatectomy: a randomized controlled trial (EORTC trial 22911).
      ] and ARO [
      • Wiegel T.
      • Bottke D.
      • Steiner U.
      • et al.
      Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pT3 prostate cancer with postoperative undetectable prostate-specific antigen: ARO 96–02/AUO AP 09/95.
      ] and 60–64 Gy for SWOG [
      • Thompson I.M.
      • Tangen C.M.
      • Paradelo J.
      • et al.
      Adjuvant radiotherapy for pathologically advanced prostate cancer: A randomized trial.
      ], or that used in the three ongoing RCT for early salvage vs. adjuvant RT trials [RAVES [12], RADICALS [
      • Parker C.
      • Clarke N.
      • Logue J.
      • et al.
      RADICALS (Radiotherapy and Androgen Deprivation in Combination after Local Surgery).
      ] and GETUG-17 [
      • Richaud P.
      • Sargos P.
      • Henriques de Figueiredo B.
      • et al.
      Postoperative radiotherapy of prostate cancer (GETUG-17).
      ]] – 64–66 Gy, or in the GETUG-16 trial of salvage RT ± ADT [
      • Carrie C.
      • Hasbini A.
      • de Laroche G.
      • et al.
      Salvage radiotherapy with or without short-term hormone therapy for rising prostate-specific antigen concentration after radical prostatectomy (GETUG-AFU 16): a randomised, multicentre, open-label phase 3 trial.
      ] – 66 Gy. Indeed, an analysis of the patterns of failure from the SWOG and EORTC adjuvant RT trials demonstrated that treatment failure is predominantly local and that therefore an improvement in local therapy (ie. presumably via radiotherapy technique and higher dose) will result in improved outcomes [
      • Thompson I.M.
      • Tangen C.M.
      • Paradelo J.
      • et al.
      Adjuvant radiotherapy for pathologically advanced prostate cancer: A randomized trial.
      ,
      • Bolla M.
      • van Poppel H.
      • Collette L.
      • et al.
      Postoperative radiotherapy after radical prostatectomy: a randomized controlled trial (EORTC trial 22911).
      ].
      The reluctance to deliver higher doses in the post-operative setting was historically based upon the belief that toxicity would be excessive and represented an unwarranted risk since no evidence for the effectiveness of higher doses existed. Similarly, until the demonstration of improved outcomes with dose-escalation for definitive RT of prostate cancer, the same beliefs prevailed and radical RT doses were universally <70 Gy until the late 1980s. Extremely little data have been published on the toxicities associated with prostate post-op RT. In a large multi-institutional analysis, radiation dose (which ranged mostly from 60 Gy to 70 Gy) was not predictive of either late GU or GI toxicity on multivariate analysis [
      • Feng M.
      • Hanlon A.L.
      • Pisansky T.M.
      • et al.
      Predictive factors for late genitourinary and gastrointestinal toxicity in patients with prostate cancer treated with adjuvant or salvage radiotherapy.
      ]. Overall, RTOG grade 2 GU and GI toxicity was 10% and 4%, respectively, grade 3 was only 1% and 0.4%, respectively, and grade 4 was only 0.2% and 0.3%, respectively. In a study from MSKCC [
      • Goenka A.
      • Magsanoc J.M.
      • Pei X.
      • et al.
      Improved toxicity profile following high-dose postprostatectomy salvage radiation therapy with intensity-modulated radiation therapy.
      ] using IMRT to deliver dose ⩾70 Gy (used among 94% of the patients) they report no difference between late GU toxicity at 16.8% grade ⩾2 vs. 3DCRT (which delivered a dose 66–70 Gy for 37% of patients and ⩾70 Gy for the rest) with 15.8% grade ⩾2, p = 0.86. With respect to late GI toxicities, the IMRT group fared better than with 3DCRT despite the fact that proportionally more got a higher dose (1.9% vs. 10.2% for late GI grade ⩾2, p = 0.02). Lastly, in an early report from a salvage RT dose-escalation trial comparing 64 Gy vs. 70 Gy [
      • Ghadjar P.
      • Hayoz S.
      • Bernhard J.
      • et al.
      Acute toxicity and quality of life after dose-intensified salvage radiation therapy for biochemically recurrent prostate cancer after prostatectomy: first results of the randomized trial SAKK 09/10.
      ] there were no statistically significant differences between acute grade 2 and 3 GU toxicities (13.0% and 0.6%, respectively, with 64 Gy vs. 16.6% and 1.7%, respectively, with 70 Gy, p = 0.2) nor similarly for acute grade 2 and 3 GI toxicities (16.0% and 0.6%, respectively, with 64 Gy vs. 15.4% and 2.3%, respectively, with 70 Gy, p = 0.8).

      Study limitations

      The limitations inherent to the current study design include: (1) population heterogeneity: systematic reviews of retrospective studies are inherently subject to population heterogeneity effects which tend to average out the relative influence of all contributing factors. Therefore one weakness of this study is that while it does identify the two most influential factors, it cannot exclude the relative importance of the others, particularly within the context of individual patient selection and management. It has been shown that dose–response for prostate cancer can appear to be steeper when derived from single series and shallower when derived from pooled analysis or meta-analysis [
      • Diez P.
      • Vogelius I.S.
      • Bentzen S.M.
      A new method for synthesizing radiation dose-response data from multiple trials applied to prostate cancer.
      ]. It is nevertheless striking that above this background of noise that these two factors remain prominent; (2) reporting bias: such bias is less likely an issue in this study by design. Counter to most meta-analyses, none of the published series used intended to compare one dose with another, nor were they motivated to report positive outcomes. Instead, by reporting single arm series they are therefore less prone to this type of bias; (3) effect of co-variate factors: without patient level data it is not mathematically possible to account for covariates using the technique of multivariate analysis. However, no factor, as quantified by the proportion present within each series, was not statistically associated with RT dose and therefore is consistent with the hypothesis that RT dose effect is independent; (4) narrow span of dose: the limited span of doses used in all SRT series limits the strength of any analysis of dose–response, but despite this drawback, the large number of series available appears to have compensated for this shortcoming; (5) unknown biologic confounders: PSA doubling time, well-known to be an independent indicator of biologic aggressiveness and associated with occult systemic disease, was not routinely available in published SRT series and therefore is not available for inclusion in this study.

      Proposed randomized clinical trial

      The hypothesis that salvage RT dose escalation would yield improved RFS could be tested with a modest clinical trial. Given an expected proportional gain in bRFS of 2% per incremental Gy, a randomized trial testing salvage RT dose of 76 Gy vs. 66 Gy would be expected to detect a 20% difference in bRFS at 5 years between the two treatment arms (corresponding to a Hazard Ratio of 1.87, assuming control bRFS at 46%). Such a trial would need a total of 212 patients in each arm (assuming a power of 0.9, a two-sided alpha of 0.05, 4 years of accrual, 6 months followup intervals and 4 years of follow-up) [
      • Schoenfeld D.
      • Richter J.
      Nomograms for calculating the number of patients needed for a clinical trial with survival as an endpoint.
      ]. Naturally, the importance of technique, including strict anatomical criteria for setup as well as daily imaging such as CBCT to reproduce such anatomy would be critical in order to deliver such doses safely.

      Conclusions

      This study provides level 2a evidence for escalated SRT dose of at least 70 Gy and probably higher. The presence of an SRT dose–response for microscopic disease after surgery supports the hypothesis that prostate cancer is inherently radio-resistant. A modest randomized trial of salvage dose escalation appears feasible.

      Conflict of interest statement

      The author does not have any conflict of interest regarding the content, treatment, drugs or technology associated with this report.

      Appendix A. Supplementary data

      References

        • Thompson I.M.
        • Tangen C.M.
        • Paradelo J.
        • et al.
        Adjuvant radiotherapy for pathologically advanced prostate cancer: A randomized trial.
        JAMA. 2006; 296: 2329-2335
        • Stephenson A.J.
        • Scardino P.T.
        • Kattan M.W.
        • et al.
        Predicting the outcome of salvage radiation therapy for recurrent prostate cancer after radical prostatectomy.
        J Clin Oncol. 2007; 25: 2035-2041
        • Stephenson A.J.
        • Shariat S.F.
        • Zelefsky M.J.
        • et al.
        Salvage radiotherapy for recurrent prostate cancer after radical prostatectomy.
        JAMA. 2004; 291: 1325-1332
        • Pollack A.
        • Zagars G.K.
        • Starkschall G.
        • et al.
        Prostate cancer radiation dose-response: results of the M. D. Anderson phase III randomized trial.
        Int J Radiat Oncol Biol Phys. 2002; 53: 1097-1105
        • Zietman A.L.
        • DeSilvio M.L.
        • Slater J.D.
        • et al.
        Comparison of conventional-dose vs high-dose conformal radiation therapy in clinically localized adenocarcinoma of the prostate: A randomized controlled trial.
        JAMA. 2005; 294: 1233-1239
        • Eade T.N.
        • Hanlon A.L.
        • Horwitz E.M.
        • et al.
        What dose of external beam radiation is high enough for prostate cancer?.
        Int J Radiat Oncol Biol Phys. 2007; 68: 682-689
        • Zelefsky M.J.
        • Sa Leibel
        • Gaudin P.B.
        • et al.
        Dose–escalation with three-dimensional conformal radiation therapy affects the outcome in prostate cancer.
        Int J Radiat Oncol Biol Phys. 1998; 41: 491-500
        • Catton C.
        • Gospodarowicz M.
        • Mui J.
        • et al.
        Clinical and biochemical outcome of conventional dose radiotherapy for localized prostate cancer.
        Can J Urol. 2002; 9: 1444-1452
        • Valicenti R.K.
        • Thompson Jr., I.
        • Albertsen P.
        • et al.
        Adjuvant and salvage radiation therapy after prostatectomy: american society for radiation oncology/american urological association guidelines.
        Int J Radiat Oncol Biol Phys. 2013; 86: 822-828
        • Bolla M.
        • van Poppel H.
        • Collette L.
        • et al.
        Postoperative radiotherapy after radical prostatectomy: a randomized controlled trial (EORTC trial 22911).
        Lancet. 2005; 366: 572-578
        • Wiegel T.
        • Bottke D.
        • Steiner U.
        • et al.
        Phase III postoperative adjuvant radiotherapy after radical prostatectomy compared with radical prostatectomy alone in pT3 prostate cancer with postoperative undetectable prostate-specific antigen: ARO 96–02/AUO AP 09/95.
        J Clin Oncol. 2009; 27: 2924-2930
      1. Trans Tasman Radiation Oncology Group (TROG 08–03) RAVES trial: Radiotherapy – Adjuvant Versus Early Salvage.

        • Parker C.
        • Clarke N.
        • Logue J.
        • et al.
        RADICALS (Radiotherapy and Androgen Deprivation in Combination after Local Surgery).
        Clin Oncol (R Coll Radiol). 2007; 19: 167-171
        • Richaud P.
        • Sargos P.
        • Henriques de Figueiredo B.
        • et al.
        Postoperative radiotherapy of prostate cancer (GETUG-17).
        Cancer Radiother. 2010; 14: 500-503
        • Ghadjar P.
        • Hayoz S.
        • Bernhard J.
        • et al.
        Acute toxicity and quality of life after dose-intensified salvage radiation therapy for biochemically recurrent prostate cancer after prostatectomy: first results of the randomized trial SAKK 09/10.
        J Clin Oncol. 2015; 33: 4158-4166
        • Moher D.
        • Liberati A.
        • Tetzlaff J.
        • Altman D.G.
        • PRISMA Group
        Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement.
        PLoS Med. 2009; 6 (10.1371): e1000097
        • Okunieff P.
        • Morgan D.
        • Niemierko A.
        • Suit H.D.
        Radiation dose-response of human tumors.
        Int J Radiat Oncol Biol Phys. 1995; 32: 1227-1237
        • King C.R.
        The timing of salvage radiotherapy after radical prostatectomy: a systematic review.
        Int J Radiat Oncol Biol Phys. 2012; 84: 104-111
        • Ohri N.
        • Dicker A.P.
        • Trabulsi E.J.
        • Showalter T.N.
        Can early implementation of salvage radiotherapy for prostate cancer improve the therapeutic ratio? A systematic review and regression meta-analysis with radiobiological modelling.
        Eur J Cancer. 2012; 48: 837-844
        • Tendulkar R.D.
        • Agarwal S.
        • Gao T.
        • et al.
        Contemporary update of a multi-institutional predictive nomogram for salvage radiotherapy after radical prostatectomy.
        J Clin Oncol. 2016; 34: 3648-3654
      2. Howick J, Chalmers I, Glasziou P, et al. Oxford Centre for Evidence-Based Medicine 2011 Levels of Evidence (OCEBM LoE Working Group) (http://www.cebm.net/wp-content/uploads/2014/06/CEBM-Levels-of-Evidence-2.1.pdf).

        • Carrie C.
        • Hasbini A.
        • de Laroche G.
        • et al.
        Salvage radiotherapy with or without short-term hormone therapy for rising prostate-specific antigen concentration after radical prostatectomy (GETUG-AFU 16): a randomised, multicentre, open-label phase 3 trial.
        Lancet Oncol. 2016; 17: 747-756
        • Feng M.
        • Hanlon A.L.
        • Pisansky T.M.
        • et al.
        Predictive factors for late genitourinary and gastrointestinal toxicity in patients with prostate cancer treated with adjuvant or salvage radiotherapy.
        In J Radiat Oncol Biol Phys. 2007; 68: 1417-1423
        • Goenka A.
        • Magsanoc J.M.
        • Pei X.
        • et al.
        Improved toxicity profile following high-dose postprostatectomy salvage radiation therapy with intensity-modulated radiation therapy.
        Eur Urol. 2011; 60: 1142-1148
        • Diez P.
        • Vogelius I.S.
        • Bentzen S.M.
        A new method for synthesizing radiation dose-response data from multiple trials applied to prostate cancer.
        Int J Radiat Oncol Biol Phys. 2010; 77: 1066-1071
        • Schoenfeld D.
        • Richter J.
        Nomograms for calculating the number of patients needed for a clinical trial with survival as an endpoint.
        Biometrics. 1982; 38: 163-170