Therapies for Clinically Localized Prostate Cancer: A Comparative Effectiveness Review

You have accessJournal of UrologyReview Article1 Apr 2021Therapies for Clinically Localized Prostate Cancer: A Comparative Effectiveness Review Timothy J. Wilt, Kristen E. Ullman, Eric J. Linskens, Roderick MacDonald, Michelle Brasure, Elizabeth Ester, Victoria A. Nelson, Jayati Saha, Shahnaz Sultan, and Philipp Dahm Timothy J. WiltTimothy J. Wilt †Correspondence: Minneapolis VA Healthcare System, Minneapolis, Minnesota E-mail Address: [email protected] Minneapolis VA Healthcare System, Minneapolis, Minnesota , Kristen E. UllmanKristen E. Ullman Minneapolis VA Healthcare System, Minneapolis, Minnesota , Eric J. LinskensEric J. Linskens Minneapolis VA Healthcare System, Minneapolis, Minnesota , Roderick MacDonaldRoderick MacDonald Minneapolis VA Healthcare System, Minneapolis, Minnesota , Michelle BrasureMichelle Brasure Minneapolis VA Healthcare System, Minneapolis, Minnesota , Elizabeth EsterElizabeth Ester Minneapolis VA Healthcare System, Minneapolis, Minnesota , Victoria A. NelsonVictoria A. Nelson Minneapolis VA Healthcare System, Minneapolis, Minnesota , Jayati SahaJayati Saha Minneapolis VA Healthcare System, Minneapolis, Minnesota , Shahnaz SultanShahnaz Sultan Minneapolis VA Healthcare System, Minneapolis, Minnesota , and Philipp DahmPhilipp Dahm Minneapolis VA Healthcare System, Minneapolis, Minnesota View All Author Informationhttps://doi.org/10.1097/JU.0000000000001578AboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack CitationsPermissionsReprints ShareFacebookTwitterLinked InEmail Abstract Purpose: We sought to identify new information evaluating clinically localized prostate cancer therapies. Materials and Methods: Bibliographic databases (2013–January 2020), ClinicalTrials.gov and systematic reviews were searched for controlled studies of treatments for clinically localized prostate cancer with duration ≥5 years for mortality and metastases, and ≥1 year for harms. Results: We identified 67 eligible references. Among patients with clinically, rather than prostate specific antigen, detected localized prostate cancer, watchful waiting may increase mortality and metastases but decreases urinary and erectile dysfunction vs radical prostatectomy. Comparative mortality effect may vary by tumor risk and age but not by race, health status, comorbidities or prostate specific antigen. Active monitoring probably results in little to no mortality difference in prostate specific antigen detected localized prostate cancer vs radical prostatectomy or external beam radiation plus androgen deprivation regardless of tumor risk. Metastases were slightly higher with active monitoring. Harms were greater with radical prostatectomy than active monitoring and mixed between external beam radiation plus androgen deprivation vs active monitoring. 3-Dimensional conformal radiation and androgen deprivation plus low dose rate brachytherapy provided small mortality reductions vs 3-dimensional conformal radiation and androgen deprivation but little to no difference on metastases. External beam radiation plus androgen deprivation vs external beam radiation alone may result in small mortality and metastasis reductions in higher risk disease but may increase sexual harms. Few new data exist on other treatments. Conclusions: Radical prostatectomy reduces mortality vs watchful waiting in clinically detected localized prostate cancer but causes more harms. Effectiveness may be limited to younger men and those with intermediate risk disease. Active monitoring results in little to no mortality difference vs radical prostatectomy or external beam radiation plus androgen deprivation. Few new data exist on other treatments. Abbreviations and Acronyms 3D-CRT 3-dimensional conformal radiation therapy ADT androgen deprivation therapy AHRQ Agency for Healthcare Research and Quality AM active monitoring AUA American Urological Association BT brachytherapy CLPC clinically localized prostate cancer COE certainty of evidence EBRT external beam radiation therapy HIFU high intensity focused ultrasound LDR-PB low dose rate prostate brachytherapy MRI magnetic resonance imaging PCa prostate cancer PDT photodynamic therapy PIVOT Prostate Cancer Intervention vs Observation Trial PSA prostate specific antigen RCT randomized controlled trial ROB risk of bias RP radical prostatectomy SPCG4 Scandinavian Prostate Cancer Group Study Number 4 WW watchful waiting In 2020, prostate cancer was estimated to be the most frequently diagnosed nondermatological malignancy (191,930 new cases) and the second leading cause of cancer death (33,330) among men in the United States.1 Treatment related medical costs were projected to rise to $16 billion per year by the end of 2020. In about 90% of men diagnosed with prostate cancer, the disease is confined to the prostate gland (clinically localized prostate cancer).2 Although disease progression can result in morbidity and mortality, most cases of clinically localized prostate cancer grow slowly and remain asymptomatic, even if untreated. Attention is turning to potentially lower risk focal therapies that focus treatment on the index lesion, such as HIFU and cryotherapy.3–5 Use of these options has also increased in response to advances in MRI technology, which allow for better detection of local lesions potentially treatable with "lesion targeted" interventions rather than whole-gland therapy. Awareness has increased regarding the slow growing nature of most prostate specific antigen detected tumors and the importance of weighing treatment benefits and harms to avoid treatment related complications.6 Thus, treatments for clinically localized prostate cancer aim to balance benefits with complications, burden and costs. The purpose of this review was to identify new information and update previous Agency for Healthcare Research and Quality and American Urological Association funded reviews7–9 evaluating treatments for CLPC. Findings can inform clinical guideline committees as they update guidelines. Materials and Methods We employed methods consistent with the AHRQ Evidence-Based Practice Center Program Methods Guidance (https://effectivehealthcare.ahrq.gov/topics/cer-methods-guide/overview). We describe these in the full report (https://effectivehealthcare.ahrq.gov/products/prostate-cancer-therapies/report). Randomized controlled trials ere assessed for risk of bias using the Cochrane ROB tool.10 The tool includes domains for random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other sources of bias. RCTs were classified as low, moderate or high ROB based on the collective ROBs across domains. We assessed observational studies using the ROBINS-1 tool.11 Observational studies were rated as low, moderate, serious or critical ROB based on the ROBIN-1 criterion. We referenced findings from the 2014 AHRQ8 and 2016 AUA9 funded reviews and included them in updated analyses if RCTs provided additional data on similar populations, interventions, comparators, and outcomes. We summarized and compared major findings from our review with those of the prior reports. We derived a priori thresholds defining "small," "moderate" and "large" effect sizes for benefits and harms (supplementary Appendix 1, https://www.jurology.com). Our searches covered publication dates from January 2013 to January 2020. We modified Grading of Recommendations Assessment, Development and Evaluation and Evidence-Based Practice Center tools for ROB and COE assessments (supplementary Appendix 2, https://www.jurology.com). We included controlled studies of CLPC (stages T1–T3a) treatments with duration ≥5 years for mortality and metastases and ≥1 year for quality of life and harms (supplementary Appendix 3, https://www.jurology.com). We extracted inclusion and exclusion criteria; sample size; participant age, race, clinical stage, Gleason score, and tumor risk classification and score; intervention and comparator characteristics; followup duration; and results for outcomes and adverse effects. We extracted data at 1 year and the longest followup for quality of life, health status, and harms; and we extracted data at 5-year intervals for mortality and metastases or at mean/median followup if that was the only way reported. One investigator extracted data to tables with verification by a second reviewer. One investigator rated ROB, extracted data and assessed COE, and a second checked accuracy. We analyzed English-language studies with low or medium ROB. We compiled results in evidence tables and synthesized evidence for each unique comparison with meta-analysis when appropriate. We assessed clinical and methodological heterogeneity to determine appropriateness of pooling data.12 When able to pool data, the Hartung-Knapp-Sidik-Jonkman method for random effects models was applied when there were at least 5 trials, and a fixed effect model was used when there were fewer than 5 trials and no between-study variance (tau2 at or near 0). When meta-analysis was not appropriate, we summarized findings. We calculated RRs or Peto ORs and absolute risk differences with the corresponding 95% CIs for binary outcomes. Mean differences and/or standardized mean differences with 95% CIs were calculated for continuous outcomes. Data were analyzed in Comprehensive Meta-Analysis™ version 3 or R software (package "meta") version 3.6.0 (R Project for Statistical Computing, Vienna, Austria). We assessed COE using the Grading of Recommendations Assessment, Development and Evaluation approach for key outcomes.13 This included assessing the applicability of results by analyzing the study population, diagnostic approaches, eligibility criteria, patient and intervention characteristics and other potential factors that may differ from current populations of treatment-naïve men with CLPC. For each comparison, one investigator rated the COE for each outcome as high, moderate, low or insufficient. COE was reviewed by a second investigator. We resolved discrepancies by consensus. Results Our search identified 11,327 references (fig. 1). Title and abstract screening eliminated 10,564 references leaving 763 references for full text review. We identified 67 eligible references, of which 17 were unique RCTs. A list of all eligible publications can be found in supplementary Appendix 4 (https://www.jurology.com). Supplemental searches of ClinicalTrials.gov and other grey literature sources did not yield additional eligible studies. Figure 2 illustrates intervention comparisons addressed in eligible RCTs according to study sample size.Table 1 summarizes findings by interventions and outcomes from the prior AHRQ and AUA funded reviews, and updated findings derived from our report. Intervention comparisons considered either "out of scope" for this review, or where we found no new data, are summarized in supplementary Appendix 5 (https://www.jurology.com). We provide a narrative summary of benefits and harms according to intervention and comparison. Additional information on effect estimates for individual studies and COE for all-cause mortality, prostate cancer mortality, and metastases are provided in tables 2–5. Effect estimates and COE for harms and quality of life data are provided in supplementary Appendix 6 (https://www.jurology.com). Figure 1. Literature flow diagram. SR, systematic review. Figure 2. Plot of comparisons addressed in RCTs identified in updated literature search. Node size reflects sample size. Width of lines reflects number of RCTs that evaluated that comparison. Within-category comparisons are not shown in figure. One RCT (ProtecT) was 3-arm trial. Active surveillance protocols varied. Prior systematic reviews identified 3 additional trials that compared external beam radiation alone with add-on androgen deprivation therapy. Table 1. Summary updates of comparisons between reviews Intervention/Comparison Outcome(s) Previous Findings from 2014 AHRQ or 2016 AUA Funded Reviews* Present Findings Derived from Studies Published after Prior Reviews and by Incorporating Prior RCT Data when Applicable† WW vs RP in men with clinically detected (SPCG4) or mainly clinically detected (PIVOT) CLPC† All-cause mortality, PCa-specific mortality, metastases harms Insufficient evidence on all-cause mortality, PCa-specific mortality, and erectile and bowel harms; RP probably reduces metastases; WW may reduce urinary harms; insufficient evidence for erectile and bowel harms‡ WW vs RP in men with clinically detected CLPC (SPCG4)—probably results in moderate increases in all-cause mortality and large increases in PCa-specific mortality and metastases at 25 yrs; mortality effects may be limited to men younger than age 65 yrs and men with intermediate risk CLPC; no new data for harms WW vs RP in men with mainly clinically detected CLPC (PIVOT)—probably results in moderate increase in all-cause mortality and large reduction in metastases, and small increase in PCa-specific mortality and at 20 yrs; mortality effects may be limited to men younger than age 65 yrs and men with intermediate risk CLPC; probably results in moderate reduction in erectile and urinary harms at 10 yrs AM (PSA based) vs EBRT+ADT All-cause mortality, PCa-specific mortality, metastases, harms Not addressed AM vs EBRT+ADT in men with PSA screen detected CLPC—probably results in little to no difference in all-cause mortality, may result in little to no difference in PCa-specific mortality and probably results in small increases in metastases at 10 yrs; results may not vary by patient or tumor characteristics; may result in small decrease in erectile dysfunction, probably results in small increase in urinary incontinence and may make little to no difference in fecal incontinence at 6 yrs AM (PSA based) vs RP All-cause mortality, PCa-specific mortality, metastases, harms Not addressed AM vs RP in men with PSA screen detected CLPC—may result in little to no difference in all-cause or PCa-specific mortality but probably results in small increase in metastases at 10 yrs; results may not vary by pt or tumor characteristics; probably results in large decrease in erectile dysfunction and moderate decrease in urinary incontinence, and may make little to no difference in fecal incontinence at 6 yrs AS (biopsy+PSA based) vs PDT Harms Not addressed AS vs PDT in men with PSA screen detected low risk CLPC—probably results in large decrease in erectile dysfunction and moderate decrease in urinary retention at 2 yrs RP vs EBRT+ADT All-cause mortality, PCa-specific mortality, metastases, harms Clinical outcomes not addressed Insufficient evidence on harms§ RP vs EBRT+ADT in men with PSA screen detected CLPC—may result in little to no difference in all-cause mortality, PCa-specific mortality and metastases at 10 yrs; results on PCa-specific mortality may not differ by age, PSA level, Gleason score or clinical stage; probably results in increase in erectile and urinary harms and decrease in bowel dysfunction at 6 yrs RP+ADT vs EBRT+HDR BT+ADT All-cause mortality, PCa-specific mortality, harms Insufficient evidence on harms for RP vs EBRT+BT§ RP+ADT vs EBRT+high dose rate BT+ADT in men with T1b–T3a PCa of any histological grade—may result in small increase in erectile dysfunction at 2 yrs; insufficient evidence on urinary or bowel harms at 2 yrs and all-cause or PCa-specific mortality through 10 yrs RP vs HIFU Harms Not addressed In men with Gleason score 7,