An Update of the European/ Swedish Screening Trials Per-Anders Abrahamsson MD, PhD, Professor Department of Urology Lund University, Malmö, Sweden Secretary General of EAU 2007–2015 27th Interna+onal Prostate Cancer Update (IPCU), Colorado, 2017 Disclosures • FERRING PharmaceuKcals A/S • CAMURUS AB, Sweden • CERNELLE AB; Sweden • GOAR HOLDING AS, Norway 3 large-scale randomized trials addressing the effecAveness of Prostate-Specific AnAgen (PSA) screening on prostate cancer mortality • Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO) • European Randomised Study of Screening for Prostate Cancer (ERSPC) • Göteborg Randomised PopulaAon-Based Prostate Cancer Screening Trial (Göteborg) Andriole GL, et al. New Engl J Med. 2009;360(13):1310-9. Schröder FH, et al. New Engl J Med. 2009;360(13):1320-8. Hugosson J, Carlsson S et al. Lancet Oncol. 2010;11:725-32. History of ERSPC • IniKated by Prof. Fritz H. Schröder (The Netherlands) in the early 1990’s 1991-1993 – Netherlands and Belgium 1995-1996 – The Göteborg trial (Sweden) led by Prof. Jonas Hugosson started as an independent trial with N=20,000 men ages 50–69; it joined ERSPC in 1996 with men in the core age group 55–69 1996 – Finland, Spain, Italy, Switzerland 2002 – France • 8 European countries • 162,388 men aged 55–69 were randomized • Study is sKll ongoing (2017), with support from EAU • Now led by Prof. Jonas Hugosson Schröder FH, Roobol MJ. Eur Urol. 2010;58(1):46-52. ERSPC trial shows 21% reducAon in cancer deaths at 13 years NNI = number needed to invite to screening to save one man from dying from prostate cancer NND = number needed to diagnose to save one man from dying from prostate cancer Schröder, et al. Lancet. 2014;384(9959):2027-35. Prostate cancer deaths reduced by ≈ 40% at more than 14 years follow-up in Göteborg trial Hugosson J, Carlsson S et al. Lancet Oncol. 2010;11:725-32. • Prostate cancer incidence: – HR 5.2 in year 1 à 1.1 during follow-up • Prostate cancer mortality: – Screening: 0.98% – Control: 1.50% HR 0.65, 95% CI 0.49-0.87 = 35% relaAve risk reducAon • Even if the relave difference in mortality is decreasing between the arms with longer follow-up, the prostate cancer incidence difference is also decreasing. • This together with an increased absolute difference in prostate cancer mortality results in more favorable outcome measures: NNI=231 and NND=10. Arnsrud Godtman R, et al. Abstract presented at EAU annual meeKng in Munich 2016. No evidence of a prostate cancer mortality benefit for organized versus opportunisc screening at 15 years in the PLCO trial Screening group Control group RR PCDx: 1.27 (95% CI, 1.20-1.35) RR PCSM: 1.04 (95% CI, 0.87-1.24) # prostate cancer deaths Years from randomizaon Pinsky P, et al. Cancer. 2016. In press. SubstanAal PSA tesAng in the control arm of PLCO •~90% of control parKcipants had at least one PSA test before/during the trial (Shoag, 2016) •86% of the men in the control arm and 99% of the men in the intervenKon arm received any PSA tesKng during the trial (15 years of follow-up) (Pinsky, 2016) •Yearly PSA tesKng rates were 46% and 84%, respecKvely (Pinsky, 2016) Shoag J et al. N Engl J Med. 2016:374;18. Pinsky P et al. Clin Trials. 2010;7:303. Andriole GL, et al. JNCI. 2012;104(2):125-32. Pinsky P, et al. Cancer. 2016 In press. PSA tesAng in Europe/Sweden was low at study start • Naonal guidelines do not recommend PSA screening • Only 3% pre-screened in the Göteborg trial • Exposure to PSA-tesKng increased over Kme; PSA uptake in men aged 55 to 69 years increased from 0 to 56% between 1997 and 2007 in Sweden in general • 40% of the cancers in the control group in Göteborg were diagnosed through opportunisKc screening in asymptomac men at 18 years of follow-up Hugosson J, et al. Lancet Oncology. 2010;11(8):725-32. Schröder F, et al. Eur Urol. 2010; 58:46-52. Jonsson H, et al. Int J Cancer. 2014;106(3):dju007. Arnsrud Godtman R, et al. Eur Urol. 2015;68(3):354-60. Organized PSA-tesAng in Göteborg reduces the risk of dying from prostate cancer but increases the number of cancers detected PROSTATE CANCER INCIDENCE Göteborg screening arm Göteborg control arm = =organized opportunisc overdiagnosis no screening no screening Arnsrud Godtman R, et al. Eur Urol. 2015;68(3):354-60. Opportunisc PSA-tesAng has li_le -- if any -- effect on prostate cancer mortality at 18 years and results in overdiagnosis PROSTATE CANCER MORTALITY No effect on no screening no screening prostate cancer mortality organized opportunisc Arnsrud Godtman R, et al. Eur Urol. 2015;68(3):354-60. Almost twice the number of men needed to be diagnosed to save one man from dying from prostate cancer with opportunisc screening compared to organized screening 18 years of Organized screening Opportunisc screening follow-up vs. no screening vs. no screening NNI 139 493 (95% CI 107 to 200) (95% CI 213 to -1563) NND 13 23 NNI = number needed to invite to screening to save one man from dying from prostate cancer NND = number needed to diagnose to save one man from dying from prostate cancer Arnsrud Godtman R, et al. Eur Urol. 2015;68(3):354-60. Possible reasons why Göteborg trial show a large reducAon in prostate cancer mortality • Younger men at start – median age 56 at baseline; 60 in ERSPC • Lower PSA-threshold for prostate biopsy – ≥ 2.5-3 ng/mL; ≥ 3-4 ng/mL in ERSPC • More intense screening – every 2 years; 4-7 years in ERSPC (although less intense as PLCO=annual) • Long duraon of screening – every 2 years from ages 50-64 up to the age of 70 (median 10 years) • High biopsy compliance – 93%; 30-40% in PLCO • Limited contaminaon • Long reported follow-up – 14-18 years Hugosson J, Carlsson S et al. Lancet Oncol. 2010;11:725-32. Schröder FH, et al. Lancet. 2014;384(9959):2027-35. • Screened group =1,756 parKcipants in the screening arm of the Göteborg trial (screened every 2 years) • Unscreened group = 1,162 unscreened men parKcipang in the Malmö PrevenKve Project in 1974-86, with PSA levels measured retrospecKvely in stored blood samples Carlsson S, et al. BMJ 2014:348:g2296. The raAo of benefits to harms of PSA screening varies noAceably with blood PSA levels at age 60 • Median PSA at age 60 is ~1.0 ng/mL • For men with PSA < 2 ng/mL (72% of the populaon) at age 60, conKnued screening lead to a large increase in prostate cancer incidence, but no decrease in mortality over 15 years of follow-up • For men with PSA ≥ 2 ng/mL at age 60, conKnued screening was associated with a large reducKon in cancer death; NNI=23 and NND=6 Carlsson S, et al. BMJ 2014:348:g2296. StarAng regular screening at ages 50-54 reduced prostate cancer mortality by 71% at 17 years compared to no screening 71% ↓ @ 17 yrs NNI 176 Unscreened NND 16 (Malmö PrevenAve Project, Sweden) Screened (Göteborg screening arm, Sweden) N=3479 screened (50-54) Prostate cancer mortality N=4060 unscreened (51-55) PSA ≥3 ng/mL cut-off for biopsy RR PCDx: 2.56 (95% CI 2.18-3.02); 15% vs. 6.3% RR PCSM: 0.29 (95% CI 0.11-0.67); 0.2% vs. 0.8% ARR PCSM: 57 fewer PC deaths / 10,000 men (95% CI 22-92) Carlsson S, et al. Eur Urol. 2017;71(1):46-52. The Stockholm-3 (STHLM3) study was innovaAve in tesAng a mulplex ,“next-generaon”, one-me, screening strategy • Stockholm, Sweden • N=145,905 men • Age 50-69 • Paired, screen- posive design • Non-inferiority • PSA ≥ 1 • 10% STHLM3 risk of high- grade PC = equal sensiKvity to PSA ≥ 3 ng/mL Grönberg H, et al. Lancet Oncol. 2015;16:1667-76. Eklund M, et al. Nat Rev Clin Oncol. 2016;13:394. Consistent with prior observaons (PHI, 4Kscore), adding kallikrein markers to the STHLM3 model improves predicAons of high-grade prostate cancer over PSA alone It is unclear whether the geneKc score adds predicKve value to the protein markers, since it was added to the stepwise model before the markers Feasibility of DRE and TRUS- volume as screening tools? No repeat screening? Carlsson S, Kaan M. Nat Rev Clin Oncol. 2016;13(3):140-2. Grönberg H, et al. Lancet Oncol. 2015;16:1667-76. Eklund M, et al. Nat Rev Clin Oncol. 2016;13:394. The STHLM3 test reduced overdiagnosis by 17% and unnecessary biopsies by 32% ?? Number of high-grade cancers caught/missed by the model compared to PSA alone not reported (due to the paired design) Grönberg H, et al. Lancet Oncol. 2015;16:1667-76. Eklund M, et al. Nat Rev Clin Oncol. 2016;13:394. Summary of an update of the European/Swedish screening trials • With longer follow-up in the trials, the effect of regular PSA- based screening on significantly reducing prostate cancer mortality is sustained • NNI and NND become favorable with longer follow-up • Screening should not start too late in life (likely no later than 55) • Regular, organized, screening may be more beneficial than opportunisKc screening in terms of net benefit • Screening can be risk-strafied, taking into account a man’s age, PSA-levels, and general health to determine who, when and how oen to screen I appreciate your aenon!!! .