Alex Tsodikov, PhD; Roman Gulati, MS; Eveline A.M. Heijnsdijk, PhD; Paul F. Pinsky, PhD; Sue M. Moss, PhD; Sheng Qiu, MS; Tiago M. de Carvalho, MS; Jonas Hugosson, MD; Christine D. Berg, MD; Anssi Auvinen, MD; Gerald L. Andriole, MD; Monique J. Roobol, PhD; E. David Crawford, MD; Vera Nelen, MD; Maciej Kwiatkowski, MD; Marco Zappa, PhD; Marcos Luján, MD; Arnauld Villers, MD; Eric J. Feuer, PhD; Harry J. de Koning, MD; Angela B. Mariotto, PhD; Ruth Etzioni, PhD
Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Cancer Institute.
Grant Support: By National Cancer Institute Award number U01CA157224.
Disclosures: Dr. Moss reports grants from the Prostate Cancer Research Foundation and the European Association of Urology during the conduct of the study. Dr. Auvinen reports personal fees from EPID Research and MSD outside the submitted work. Dr. Kwiatkowski reports personal fees from Myriad, Astellas, and Janssen outside the submitted work. Dr. Luján reports a grant from Fondo de Investigación Sanitaria during the conduct of the study. Dr. Etzioni reports personal fees from GRAIL outside the submitted work and ownership of equity in Seno. Authors not named here have disclosed no conflicts of interest. Disclosures can also be viewed at www.acponline.org/authors/icmje/ConflictOfInterestForms.do?msNum=M16-2586.
Editors' Disclosures: Christine Laine, MD, MPH, Editor in Chief, reports that she has no financial relationships or interests to disclose. Darren B. Taichman, MD, PhD, Executive Deputy Editor, reports that he has no financial relationships or interests to disclose. Cynthia D. Mulrow, MD, MSc, Senior Deputy Editor, reports that she has no relationships or interests to disclose. Deborah Cotton, MD, MPH, Deputy Editor, reports that she has no financial relationships or interest to disclose. Jaya K. Rao, MD, MHS, Deputy Editor, reports that she has stock holdings/options in Eli Lilly and Pfizer. Sankey V. Williams, MD, Deputy Editor, reports that he has no financial relationships or interests to disclose. Catharine B. Stack, PhD, MS, Deputy Editor for Statistics, reports that she has stock holdings in Pfizer and Johnson & Johnson.
Reproducible Research Statement:Study protocol: Not available. Statistical code: Source code or runs using the FHCRC model are available from Mr. Gulati (e-mail, email@example.com), runs using the MISCAN model are available from Dr. Heijnsdijk (e-mail, firstname.lastname@example.org), and source code or runs using the UMICH model are available from Dr. Tsodikov (e-mail, email@example.com). Data set: PLCO data are available from the National Cancer Institute Cancer Data Access System (https://biometry.nci.nih.gov/cdas). ERSPC data may be available from Dr. Moss (e-mail, firstname.lastname@example.org).
Requests for Single Reprints: Ruth Etzioni, PhD, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, M2-B230, PO Box 19024, Seattle, WA 98109; e-mail, email@example.com.
Current Author Addresses: Dr. Tsodikov and Mr. Qiu: Department of Biostatistics, University of Michigan, 1415 Washington Heights, Ann Arbor, MI 48109-2029.
Mr. Gulati and Dr. Etzioni: Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, 1100 Fairview Avenue North, Seattle, WA 98109-1024.
Drs. Heijnsdijk and de Koning and Mr. de Carvalho: Department of Public Health, Erasmus Medical Center, PO Box 2040, 3000 CA Rotterdam, the Netherlands.
Dr. Pinsky: Division of Cancer Prevention, National Cancer Institute, 9609 Medical Center Drive, Bethesda, MD 20892.
Dr. Moss: Wolfson Institute, Queen Mary University of London, Charterhouse Square, London EC1M 6BQ, United Kingdom.
Dr. Hugosson: Department of Urology, Sahlgrenska University Hospital, Blå stråket 5, 413 45 Göteborg, Sweden.
Dr. Berg: Department of Radiation Oncology and Molecular Radiation Sciences, Johns Hopkins Medicine, 401 North Broadway, Baltimore, MD 21231.
Dr. Auvinen: School of Health Sciences, PL 100 33014 University of Tampere, Finland.
Dr. Andriole: Division of Urologic Surgery, Department of Surgery, Washington University School of Medicine, 4921 Parkview Place, St. Louis, MO 63110.
Dr. Roobol: Department of Urology, Erasmus Medical Center, PO Box 2040, 3000 CA Rotterdam, the Netherlands.
Dr. Crawford: Urologic Oncology, University of Colorado, 1665 Aurora Court, Aurora, CO 80045.
Dr. Nelen: Provinciaal Instituut voor Hygiëne, Kronenburgstraat 45, 2000 Antwerpen, Belgium.
Dr. Kwiatkowski: Department of Urology, Kantonsspital Aarau, CH5001 Aarau, Switzerland.
Dr. Zappa: Unit of Epidemiology, Institute for Cancer Prevention, Via delle Oblate 2, 50141, Florence, Italy.
Dr. Luján: Servicio de Urología, Hospital Universitario Infanta Cristina, Universidad Complutense de Madrid, Avenida Nueve de Junio n° 2, Parla, 28981 Madrid, Spain.
Dr. Villers: Department of Urology, CHU Lille, Université de Lille, F-59000 Lille, France.
Drs. Feuer and Mariotto: Division of Cancer Control and Population Sciences, National Cancer Institute, 9609 Medical Center Drive, Bethesda, MD 20892.
Author Contributions: Conception and design: A. Tsodikov, R. Gulati, J. Hugosson, E.D. Crawford, A. Villers, E.J. Feuer, H.J. de Koning, R. Etzioni.
Analysis and interpretation of the data: A. Tsodikov, R. Gulati, E.A.M. Heijnsdijk, P.F. Pinsky, S. Qiu, C.D. Berg, A. Auvinen, G.L. Andriole, M. Zappa, A. Villers, H.J. de Koning, A.B. Mariotto, R. Etzioni.
Drafting of the article: A. Tsodikov, R. Gulati, E.D. Crawford, H.J. de Koning, R. Etzioni.
Critical revision of the article for important intellectual content: A. Tsodikov, R. Gulati, E.A.M. Heijnsdijk, P.F. Pinsky, S.M. Moss, T.M. de Carvalho, C.D. Berg, A. Auvinen, G.L. Andriole, M.J. Roobol, M. Kwiatkowski, A. Villers, E.J. Feuer, H.J. de Koning, R. Etzioni.
Final approval of the article: A. Tsodikov, R. Gulati, E.A.M. Heijnsdijk, P.F. Pinsky, S.M. Moss, S. Qiu, T.M. de Carvalho, J. Hugosson, C.D. Berg, A. Auvinen, G.L. Andriole, M.J. Roobol, E.D. Crawford, V. Nelen, M. Kwiatkowski, M. Zappa, M. Luján, A. Villers, E.J. Feuer, H.J. de Koning, A.B. Mariotto, R. Etzioni.
Provision of study materials or patients: J. Hugosson, G.L. Andriole, E.D. Crawford, M. Luján, A. Villers, H.J. de Koning.
Statistical expertise: A. Tsodikov, R. Gulati, P.F. Pinsky, S. Qiu, E.J. Feuer, R. Etzioni.
Obtaining of funding: A. Tsodikov, G.L. Andriole, M. Luján, A. Villers, H.J. de Koning, A.B. Mariotto, R. Etzioni.
Administrative, technical, or logistic support: A. Tsodikov, R. Gulati, C.D. Berg, E.J. Feuer, A.B. Mariotto.
Collection and assembly of data: A. Tsodikov, R. Gulati, P.F. Pinsky, S.M. Moss, S. Qiu, J. Hugosson, C.D. Berg, A. Auvinen, G.L. Andriole, M.J. Roobol, E.D. Crawford, V. Nelen, M. Kwiatkowski, H.J. de Koning, A.B. Mariotto.
The ERSPC (European Randomized Study of Screening for Prostate Cancer) found that screening reduced prostate cancer mortality, but the PLCO (Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial) found no reduction.
To evaluate whether effects of screening on prostate cancer mortality relative to no screening differed between the ERSPC and PLCO.
Cox regression of prostate cancer death in each trial group, adjusted for age and trial. Extended analyses accounted for increased incidence due to screening and diagnostic work-up in each group via mean lead times (MLTs), which were estimated empirically and using analytic or microsimulation models.
Randomized controlled trials in Europe and the United States.
Men aged 55 to 69 (ERSPC) or 55 to 74 (PLCO) years at randomization.
Prostate cancer screening.
Prostate cancer incidence and survival from randomization; prostate cancer incidence in the United States before screening began.
Estimated MLTs were similar in the ERSPC and PLCO intervention groups but were longer in the PLCO control group than the ERSPC control group. Extended analyses found no evidence that effects of screening differed between trials (P = 0.37 to 0.47 [range across MLT estimation approaches]) but strong evidence that benefit increased with MLT (P = 0.0027 to 0.0032). Screening was estimated to confer a 7% to 9% reduction in the risk for prostate cancer death per year of MLT. This translated into estimates of 25% to 31% and 27% to 32% lower risk for prostate cancer death with screening as performed in the ERSPC and PLCO intervention groups, respectively, compared with no screening.
The MLT is a simple metric of screening and diagnostic work-up.
After differences in implementation and settings are accounted for, the ERSPC and PLCO provide compatible evidence that screening reduces prostate cancer mortality.
National Cancer Institute.
Table 1. Summary of Participant Characteristics, Follow-up, and Prostate Cancer Cases and Deaths in the ERSPC and PLCO, Under All Available Follow-up and Restricted to 11 Years of Follow-up
Estimated MLTs in the intervention and control groups of the ERSPC and PLCO relative to a hypothetical no-screening setting (where MLT equals zero).
Estimated MLTs are visualized as increasing to the left to suggest the extent to which prostate cancer diagnosis is advanced by more intensive screening and diagnostic work-up. ERSPC = European Randomized Study of Screening for Prostate Cancer; FHCRC = Fred Hutchinson Cancer Research Center; MISCAN = Erasmus University Medical Center MIcrosimulation SCreening ANalysis; MLT = mean lead time; PLCO = Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial; UMICH = University of Michigan.
Table 2. Results of Traditional and Extended Cox Regression Analyses of Death From Prostate Cancer and Estimated Mortality Reductions in the ERSPC and PLCO Intervention Groups Relative to No Screening
Prostate cancer survival from randomization in the ERSPC and PLCO, estimated by Kaplan–Meier or Cox regression model using mean lead time estimated with the empirical approach.
ERSPC = European Randomized Study of Screening for Prostate Cancer; PLCO = Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial.
The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.
Philippe Autier, MD, Peter Boyle, PhD, F Med Sci
University of Strathclyde
September 11, 2017
Tsodikov and colleagues (1) present a complex re-analysis of the PLCO and the ERSPC trials based on screening intensity. Screening intensity in each of the four randomization groups was estimated using the mean lead time (MLT) that quantifies by how long the diagnosis of prostate cancer has been advanced. The variable “randomization group” in Cox models was then replaced by the value of the MLT found for each randomization group. In our opinion, this statistical approach is incorrect.We are not aware of any other trial report in which the randomization group variable is replaced by a variable representing a parameter calculated on the minority of subjects diagnosed with the disease in question. This replacement maneuver equates to assigning to subjects never affected by prostate cancer a statistical quantity that is entirely specific of subjects diagnosed with prostate cancer. A strange consequence is the assignment of a time variable of at least 1.6 years to those subjects who were no longer in the trial after the first 1.6 years, e.g., because of death.Overdiagnosis exerts a substantial influence on the results displayed in Table 2 of Tsodikov and colleagues (1). The MLT is a mix of clinical lead time associated with the early detection of potentially deadly cancer and of length time associated with the detection of cancer that would never have been appeared clinically during subject’s lifetime. Overdiagnosis does not affect the risk of prostate cancer death, but the amount of overdiagnosis is highly dependent on the way screening is performed. Restrained overdiagnosis would have led to MLTs shorter than 4 years in groups invited to screening, while more overdiagnosis would have led to MLTs longer than 4 years. The direct consequence of variable amounts of overdiagnosis is that hazard rates derived from Cox models could take a wide range of values, from nearly no reduction in risk of prostate cancer death to considerable risk reductions. A paradoxical consequence is that the greater the level of overdiagnosis, and thus the longer the MLT in screening groups, the smaller the apparent risk of prostate cancer death obtained from Cox models (i.e., an HR>0.92 in Table 2).Authors must provide sound methodological justification for using the replacement maneuver and require to demonstrate the absence of influence of overdiagnosis on their findings. Until then, the methods used by Tsodikov et al. should are not credible alternatives for the analysis of randomized trials of cancer screening. References1. Tsodikov A, Gulati R, Heijnsdijk EAM, Pinsky PF, Moss SM, Qiu S, et al. Reconciling the Effects of Screening on Prostate Cancer Mortality in the ERSPC and PLCO Trials. Ann Intern Med. 2017.
Alex Tsodikov, Roman Gulati, Ruth Etzioni
University of Michigan, Fred Hutchinson Cancer Research Center, Fred Hutchinson Cancer Research Center
September 12, 2017
Author's Response to Drs. Autier and Boyle
The comment by Drs. Autier and Boyle reflects an incorrect interpretation of our approach.Our approach is analogous to an analysis of a dose-response experiment, where different groups are given higher or lower doses of a treatment. The analysis could be conducted either by comparing responses between groups assigned to higher versus lower doses, or by evaluating responses given the actual doses received in each group. Given the variable amounts of screening and diagnostic work-up in each arm of the ERSPC and PLCO, we conducted the second type of analysis. In particular, the actual dose received (intensity of screening) was unknown, so we estimated it.The intensity of screening in the trials was estimated as the mean lead time (MLT), which is zero in an unscreened population and gets larger when more screening and diagnosis takes place. Importantly, it is a group-level characteristic based on the incidence in the group as a whole and is not meant to be interpreted at the individual level. Evaluating the association between the risk of prostate cancer death and this group-level characteristic is similar to the dose-response example, where the value for a given trial arm represents the average dose received among all subjects in that group.Two consequences of our approach are immediately apparent. First, because we do not propose the MLT as an individual-level quantity, the interpretation issues suggested by Drs. Autier and Boyle that arise when considering specific MLTs for specific individuals do not apply. Second, because the MLT is never based on a subset of the data within a group, it is not subject to the selection biases of as-treated analyses, which assign subjects within randomized groups values of the intervention variable based on what actually happened to them after randomization.Finally, we address the overdiagnosis issue. It is very likely that the calculated MLTs will increase if screening leads to overdiagnosis. However, if screening only leads to overdiagnosis, or if screening leads to early detection without any benefit, then the regression analysis should not show any association between the MLT and prostate cancer mortality. Our analysis did show such an association. And although it may appear small (HR=0.91–0.93 in Table 2), this is the estimated effect per year of mean lead time. Combining this estimated effect with the estimated MLT in the intervention arms of the ERSPC and PLCO trials produces the reported 25% to 32% lower risk for prostate death relative to no screening.
Anthony B. Miller, MD, Philip C. Prorok, PhD
University of Toronto, NCI
September 21, 2017
The Editor Annals of Internal Medicine Re: Reconciling the Effects of Screening on Prostate Cancer Mortality in the ERSPC and PLCO Trials This paper by Tsodikov et al (1) has a major flaw: no account is taken of the overdiagnosis induced by PSA testing. The paper introduces thea new and non-validated metric, mean lead time (MLT), defineddesigned to encompass the population diagnostic process. Since some of the cancers are screen detected, MLT is confounded with the well- known screening biases, of which overdiagnosis is an important contributor. Without adjustment for overdiagnosis (a clear harm), MLT is a flawed metric for determining screening benefit and cannot substitute for the definitive endpoint of screening, mortality from the disease. There is no indication in the paper that such an adjustment was performed. The mortality endpoint is encapsulated in the paper by Pinsky et al (2), which documents the lack of benefit from organised screening as performed in the intervention arm of the PLCO trial, compared to the opportunistic screening which occurred in the control arm. Further, the paper of Tsodikov et al (1) fails to take note ofignores the remarkablemajor discrepancies between the different components of the ERSPC trial, which have never been satisfactorily explained (3). Of particular note, the authors did not use account for treatment data in their analysis. Screening can only reduce mortality if at least some cancers are detected early and their outcome is improved due to being treatedtreatment at an earlier time than without screening. Screening benefit and treatment are inextricably connected. Whatever the true relationship between PSA screening and prostate cancer mortality reduction, it cannot be accurately determined using MLT because the definition of MLT does not include cancer characteristics, nor treatment information, and the interrelationship between the two. Thus Tsodikov, et al (1) could not adjust for the treatment imbalance reported in ERSPC (3) and so have likely overestimated any ERSPC benefit. Thus, their conclusion that the ERSPC and PLCO trials provide comparable evidence that PSA screening reduces mortality is not justified by their analysis. It is important that no one is misled by Tsodikov et al (1)In summary, their paper provides no reason to ignore the recommendation of the US Preventive Services Task Force (20124) against PSA-based screening for prostate cancer in all age groupsassume benefits in the PLCO trial, much less that the benefits are equivalent to those averred in the ERSPC trial. References 1. Tsodikov A, Gulati R, Heijnsdijk EAM, Pinsky PF, Moss SM, Qiu S, et al. Reconciling the Effects of Screening on Prostate Cancer Mortality in the ERSPC and PLCO Trials. Ann Intern Med. 2017; doi:10.7326/M16-2586 2. Pinsky PF, Prorok PC, Yu K, Kramer BS, Black A, Gohagan JK, et al. Extended mortality results for prostate cancer screening in the PLCO trial with median follow-up of 15 years. Cancer. 2017;123:592-9. [PMID: 27911486] doi:10.1002/cncr.30474 3. Miller AB. Prostate Cancer Screening. In Miller AB (Ed.) Epidemiologic Studies in Cancer Prevention and Screening. Statistics for Biology and Health Volume 79, New York, Springer Verlag, 2013 pp 277-285 Philip C. Prorok, PhD. Division of Cancer Prevention, National Cancer Institute, Bethesda, Md. Anthony B. Miller, MD. Dalla Lana School of Public Health, University of Toronto, Canada.
Stuart G. Baker
National Cancer Institute
September 28, 2017
Comment on "Reconciling the Effects of Screening on Prostate Cancer Mortality in the ERSPC and PLCO Trials"
TO THE EDITOR: The results of Tsodikov et al (1) on the effect of prostate-specific antigen (PSA) screening on prostate cancer mortality may give a distorted view of the evidence. The standard method for the analysis of clinical trials is to analyze by intent-to-treat because it uses randomization to reduce bias from unmeasured confounders. Based on data from their Table 1, the effect of screening on prostate cancer mortality per 10,000 persons was -9 with 95% CI (-15, -3) in ERSPC (European Randomized Study of Screening for Prostate Cancer) and 1 with 95% CI of (-7, 8) in the PLCO (Prostate Lung Colorectal and Ovarian Cancer Screening) trial. These results suggest a small prostate cancer mortality reduction in ERSPC and no clear evidence of a prostate cancer mortality reduction in the PLCO trial. Complicating the interpretation of the PLCO results is (i) routine PSA testing in the control arm starting at 33% and increasing to 46% in five years (2), which biases the estimated prostate cancer mortality effect towards zero, and (ii) a follow-up period after the last screen, which increases variability. Complicating the interpretation of the ERSPC results is a stage-specific imbalance in therapy between randomization groups (3) that is larger than with PLCO (4), suggesting a possible bias in treatment choice. In their combined analysis of ERPSC and PLCO, Tsodikov et al. (1) replaced randomization group by the mean lead time (MLT), a procedure which can lead to bias. Bias can arise because (i) persons screened contribute more to MLT than those not screened, (ii) the PLCO trial had more control arm screening than ERSPC, and (iii) the background rate of prostate cancer mortality (in the absence of screening) may have differed among controls screened versus not screened, analogous to substantially different rates of breast cancer mortality among refusers and controls in a mammography trial (5). Bias can also arise because (i) MLT depends on the PSA threshold, and (ii) ERSPC had a lower PSA threshold for biopsy than the PLCO trial. Consequently, ERSPC was more likely than the PLCO trial to detect overdiagnosed prostate cancers that increase MLT without contributing to prostate cancer mortality reduction. Also, Tsodikov et al. (1) reported the risk reduction in prostate cancer rather than a difference in prostate mortality rates. Reporting relative reduction rather than absolute reduction makes it difficult to weigh harms and benefits and communicate effectively to patients. Stuart G. Baker Sc.D.National Cancer InstituteBethesda, MarylandReferences1. Tsodikov A, Gulati R, Heijnsdijk EAM, et al. Reconciling the effects of screening on prostate cancer mortality in the ERSPC and PLCO trials. Ann Intern Med 2017. [Epub ahead of print]2. Pinsky PF, Blacka A, Kramer BS, Miller A, Prorok PC, Berg C. Assessing contamination and compliance in the prostate component of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. Clin Trials. 2010 ;7(4):303-11.3. Wolters T, Roobol MJ, Steyerberg EW, et al. The effect of study arm on prostate cancer treatment in the large screening trial ERSPC. Int J Cancer. 2010 ;126(10):2387-2393.4. Andriole GL, Crawford ED, Grubb RL 3rd, et al. Prostate cancer screening in the randomized Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial: mortality results after 13 years of follow-up. J Natl Cancer Inst. 2012;104(2):125-132.5. Freedman DA, Petitti DB, Robins JM. On the efficacy of screening for breast cancer. Int J Epidemiol. 2004;33(1):43-55.
Join Y. Luh, MD, Steven E. Finkelstein, MD, Jeff M. Michalski, MD, MBA, Howard Sandler MD, MS
Providence St. Joseph Health, Translational Genomics Institute, Washington University St. Louis, Cedar Sinai Medical Center
October 5, 2017
Leveling the Playing Field between PLCO and ERSPC
Screening recommendations from the USPSTF in 2012 (1) were based on two large randomized trials published in 2009 (2, 3), the European Randomized study of Screening for Prostate Cancer (ERSPC) and the Prostate, Lung, Colorectal and Ovarian (PLCO) prostate cancer screening trial, and did not include the PLCO update published in 2012. Both trials suffered from contamination in the control groups, but the worse was the PLCO trial (3,4) where with only 7 years of follow up, 85% of patients in the screening arm actually got PSA screening, and 52% of men in the control arm were screened by the 6th year. For a study with such short follow up, without scrutinizing whether or not patients were even treated, it is not surprising that no differences in prostate cancer mortality were found. Even the authors of the PLCO study stated in a later publication (5) that the PLCO study should not be interpreted as a trial of screening versus no screening, but rather as a trial of annual screening versus "usual care," which at the time the study was initiated, PSA screening was ubiquitous.In 2012, a post-hoc analysis of men in the PLCO trial (6) with little to no comorbidity showed a 44% reduction in the risk of prostate cancer mortality among men in the screening arm (hazard ratio 0.56, 95% CI 0.33-0.95), despite the contamination in the usual care group. The median age of men in this group was 61. Interestingly, the NNT (number needed to treat) to prevent one prostate cancer death at 10 years in this subgroup was only five. This analysis showed that selective use of PSA screening for men in good health appears to reduce the risk of PCSM (prostate cancer specific mortality) with minimal overtreatment.We therefore read with interest, the current study (7) that sought to remove the contamination found in the PLCO's control group and actually compare screening versus no screening. Despite the limitations in mean lead time (MLT) metrics, Dr. Tsodikov and colleagues have demonstrated that prostate screening reduces prostate cancer mortality compared to no screening, and used PLCO data to place it on a more similar playing field as the ERSPC trial.Individualized screening; taking into consideration a patient's preferences, risk factors, and the decision on whether or not to confront the risks of over-diagnosis, provides a balanced approach that will undoubtedly increase the yield in selected patients who choose to get screened. Once a diagnosis is made, multidisciplinary consultation (8) can give patients the knowledge and tools to make the best informed management decisions.1. Moyer VA; U.S. Preventive Services Task Force. Screening for prostate cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2012 Jul 17;157(2):120-34.2. Schroder FH, Hugosson J, Roobol MJ, Tammela TL, Ciatto S, Nelen V, et al; ERSPC Investigators. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med. 2009;360:1320-8.3. Andriole GL, Crawford ED, Grubb RL 3rd, Buys SS, Chia D, Church TR, et al; PLCO Project Team. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med. 2009;360:1310-9.4. Andriole GL, Crawford ED, Grubb RL 3rd,et al. Prostate cancer screening in the randomized Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial: mortality results after 13 years of follow-up. J Natl Cancer Inst. 2012 Jan 18;104(2):125-32.5. Pinsky P, Black A, Kramer BS, et al. Assessing contamination and compliance in the prostate component of the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial. Clin Trials. 2010;7 (4): 303–11.6. Crawford ED, Grubb R 3rd, Black A, et al. Comorbidity and mortality results from a randomized prostate cancer screening trial. J Clin Oncol. 2011 Feb 1;29(4):355-61.7. Tsodikov A, Gulati R, Hiejnsdijk EAM, et al. Reconciling the Effects of Screening on Prostate Cancer Mortality in the ERSPC and PLCO Trials. Ann Intern Med. 2017 Sept 5 (e-pub ahead of print).8. Aizer AA, Paly JJ, Zietman AL, et al. Multidisciplinary care and pursuit of active surveillance in low-risk prostate cancer. J Clin Oncol. 2012 Sep 1;30(25):3071-6.
October 9, 2017
Author's Response to Drs. Miller, Prorok, and Baker
Drs. Miller, Prorok, and Baker raise similar concerns about potential bias in our analysis of prostate cancer screening in the ERSPC and PLCO.Our analysis begins by quantifying the intensity of screening and diagnosis on each arm of the two trials using the mean lead time (MLT). The MLT reflects the excess incidence relative to what would be expected in the absence of screening. As noted in the comments of Drs. Miller and Prorok, and Dr. Baker, the excess incidence as quantified by the MLT includes overdiagnosis. However, this does not invalidate the MLT as a measure of screening and diagnostic intensity. We do not, as Drs. Miller and Prorok suggest, propose the MLT as a "metric for determining screening benefit" and we certainly do not propose to substitute it for the "definitive endpoint of screening, mortality from the disease." Rather, we use a well-established statistical framework to assess whether the MLT correlates with prostate cancer mortality across the study arms and whether there is evidence that any correlation observed differs across the trials. The answer to the first is yes and the answer to the second is no. These results imply that screening is beneficial relative to no screening and there is no evidence to suggest a difference in benefit between the trials. Contrary to Dr. Miller and Prorok’s suggestion, this conclusion is not at odds with the PLCO investigators who documented the lack of benefit from organized screening relative to opportunistic screening. It is true that our analysis does not explicitly account for treatments received in the trials and any clear imbalances could impact the interpretation of our results. However it appears that differences in the frequencies of curative treatment between ERSPC intervention and control arms are minor after accounting for differences in tumor stage across arms (Wolters et al., IJC, 2010). In addition, systematic differences in treatments between the ERSPC and PLCO are indirectly captured by the trial indicator in our model of prostate cancer survival, which was found to be statistically and clinically significant.Finally, pooling data from both trials, we report the relative reduction in the risk for prostate cancer death. Dr. Baker notes that this measure is less useful than the absolute reduction for communication with patients. Both relative and absolute reductions in mortality associated with screening are important for developing sound screening policies. We certainly support further work to estimate absolute reductions over time, particularly reductions that account for individual patient risk factors, as highlighted by Drs. Luh, Finkelstein, Michalski, and Sandler.ReferencesWolters T, Roobol MJ, Steyerberg EW, van den Bergh RC, Bangma CH, Hugosson J, Ciatto S, Kwiatkowski M, Villers A, Luján M, Nelen V, Tammela TL, Schröder FH. The effect of study arm on prostate cancer treatment in the large screening trial ERSPC. Int J Cancer. 2010 May 15;126(10):2387-93.
Shinkan Tokudome, MD1, Ryosuke Ando, MD2, Shuji Hashimoto, PhD3
Nisshin, 470-0114, Japan
November 2, 2017
PSA screening and prostate cancer mortality reduction: where do we really stand?
Recently, Tsodikov et al. (1) seemed to have resolved a long-standing controversy about the effectiveness of prostate-specific antigen (PSA)-based screening for mortality reduction in prostate cancer (PCa)－namely, the discrepancy between the ERSPC (European Randomized Study of Screening for Prostate Cancer) (2) and PLCO (Prostate, Lung, Colorectal, and Ovarian) Cancer Screening Trial (3). They did this by applying the same analysis method (Cox regression analysis) to mean lead times (MLTs). PCa mortality reduction in the ERSPC population with screening translated to an estimated 25%-31% compared with the population with no screening, and to 27%-32% in the PLCO population with screening. Insufficient statistical power of the ERSPC was salvaged by extending the observation period up to 11 years to increase the number of deaths in the analysis. However, there is still some debate as to whether the alleged contamination bias in the PLCO was adequately controlled for, because systematic error can be remedied only before the study is launched. Namely, there may be discussion about whether the sophisticated analysis of MLTs has been validated against a conventional survival analysis. Adjusting for major confounding factors including trial site and age, Tsodikov et al. noted longer MLTs in the PLCO than in the ERSPC. In light of this, they are advised to employ an age-specific analysis involving the elderly, because the United States Preventive Services Task Force (USPSTF) does not recommend PSA screening based on a traditional analysis for subjects aged ≥ 70 years, which is different from that for men aged 55-69 years, according to the discrepancy in benefit-harm assessments (4). The effectiveness of PCa mortality reduction by PCa risk is of importance. The authors should demonstrate whether their findings are the case not only for low-risk PCa but also for high-risk cancer, because PCa is a leading cause of cancer deaths in developed countries, and PSA testing is fraught with serious problems, including over-diagnosis and subsequent over-treatment of low-grade cancer.References1. Tsodikov A, Gulati R, Heijnsdijk EAM, Pinsky PF, Moss SE, Qiu S. Reconciling the effects of screening on prostate cancer mortality in the ERSPC and PLCO trials. Ann Intern Med. 2017;167:449-55.2. Schröder FH, Hugosson J, Roobol MJ, Tammela TL, Ciatto S, Nelen V, et al; ERSPC Investigators. Screening and prostate-cancer mortality in a randomized European study. N Engl J Med. 2009;360:1320-8.3. Andriole GL, Crawford ED, Grubb RL 3rd, Buys SS, Chia D, ChurchTR, et al; PLCO Project Team. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med. 2009;360:1310-9.4. Bibbins-Domingo K, Grossman DC, Curry SJ. The US Preventive Services Task Force 2017 Draft Recommendation Statement on Screening for Prostate Cancer: an invitation to review and comment. JAMA. 2017;317:1949-50. The authors have no financial support or conflicts of interest to declare.
Tsodikov A, Gulati R, Heijnsdijk EA, Pinsky PF, Moss SM, Qiu S, et al. Reconciling the Effects of Screening on Prostate Cancer Mortality in the ERSPC and PLCO Trials. Ann Intern Med. 2017;167:449–455. doi: 10.7326/M16-2586
Download citation file:
Published: Ann Intern Med. 2017;167(7):449-455.
Published at www.annals.org on 5 September 2017
Hematology/Oncology, Prostate Cancer.
Results provided by:
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use