Doctors are changing their recommendations: In a September, 2012, video Dr Myers went as far as to say “There is no basis for using surgery to treat Gleason 6 PCa” or Gleason 7 if there is only a small area of Gleason 4 (remember a Gleason score is the total of the two most active tissues viewed by your pathologist – so the score is out of 10.

They are obliged to report the Gleason 4 tissue even if it is tiny in volume. You can access this latest video announcement from the Ask Dr Myers site: PCa: Treating Young Men With Curative Intent 

A/prof Phillip Stricker, St Vincents Hospital announced a new procedure called Nanoknife Focal Treatment, please click on: http://www.abc.net.au/news/2013-04-14/new-pain-free-treatment-for-prostate-cancer/4624104. The disturbing thing is the hypocrisy of opposing MRI and ultrasound as diagnostic tools for years as alternatives to biopsies and now claiming that you can identify all the cancer and its location with MRI and ultrasound beggars belief. For years we have been reading about the lack of accuracy of biopsies – here is a professor, quoted in the same ABC news item saying the same:

But Professor Mark Frydenberg from Monash University says the use of focal therapy is very much still at the trial stage.

“What’s crucial is the accuracy of the biopsy in detecting the cancer,” he said.

“Through the MRI, you need a good map of the prostate. The concern is that you will end up treating disease that is not the main area of cancer, as MRIs are good but not perfect,” he said.

Sticking wires into your gland has the same inaccuracies, inflammation, infection risk, as biopsies. German hyperthermia heats the whole gland, without holes.

Remember in 1935 Upton Sinclair summed up the problem you are facing very neatly when he wrote “It is difficult to get a man (doctor) to understand something when his salary depends on not understanding it.”

T he New England Journal of  Medicine 13/7/17

Follow-up of Prostatectomy versus Observation for Early Prostate  Cancer

Timothy J. Wilt, M.D., M.P.H., Karen M. Jones, M.S., Michael J. Barry, M.D., Gerald L. Andriole, M.D., Daniel Culkin, M.D., Thomas Wheeler, M.D., William J. Aronson, M.D., and Michael K. Brawer, M.D.

From the Minneapolis Veterans Affairs (VA) Health Care System and Center for Chronic Disease Outcomes Research and the Section of General Medicine, Univer- sity of Minnesota School of Medicine, Minneapolis (T.J.W.); the VA Cooperative Studies Program Coordinating Center, Perry Point, MD (K.M.J.); the General Medicine Division, Massachusetts Gen- eral Hospital, Boston (M.J.B.); the Divi- sion of Urology, Washington University School of Medicine, St.  Louis  (G.L.A.);  the Department of Urology, University of Oklahoma College of Medicine, Oklahoma City (D.C.); the Department of Pathology and Immunology, Baylor College of Medi- cine, Houston (T.W.); VA Medical Center, Greater Los Angeles Healthcare System, Los Angeles (W.J.A.); and Myriad Genet- ics Laboratories, Salt Lake City (M.K.B.). Address reprint requests to Dr. Wilt  at  the Minneapolis VA Center for Chronic Disease Outcomes Research, 1 Veterans Dr. (111-0), Minneapolis, MN 55417, or at [email protected].

N Engl J Med 2017;377:132-42. DOI: 10.1056/NEJMoa1615869

Copyright © 2017 Massachusetts Medical Society.

BACKGROUND

We previously found no significant differences in mortality between men who under- went surgery for localized prostate cancer and those who were treated with observation only. Uncertainty persists regarding nonfatal health outcomes and long-term mortality.

METHODS

From November 1994 through January 2002, we randomly assigned 731 men with local- ized prostate cancer to radical prostatectomy or observation. We extended follow-up through August 2014 for our primary outcome, all-cause mortality, and the main sec- ondary outcome, prostate-cancer mortality. We describe disease progression, treatments received, and patient-reported outcomes through January 2010 (original follow-up).

RESULTS

During 19.5 years of follow-up (median, 12.7 years), death occurred in 223 of 364 men (61.3%) assigned to surgery and in 245 of 367 (66.8%) assigned to observation (absolute difference in risk, 5.5 percentage points; 95% confidence interval [CI], −1.5 to 12.4; hazard ratio, 0.84; 95% CI, 0.70 to 1.01; P=0.06). Death attributed to prostate cancer or treatment occurred in 27 men (7.4%) assigned to surgery and in 42 men (11.4%) as- signed to observation (absolute difference in risk, 4.0 percentage points; 95% CI, −0.2 to 8.3; hazard ratio, 0.63; 95% CI, 0.39 to 1.02; P=0.06). Surgery may have been associ- ated with lower all-cause mortality than observation among men with intermediate-risk disease (absolute difference, 14.5 percentage points; 95% CI, 2.8 to 25.6) but not among those with low-risk disease (absolute difference, 0.7 percentage points; 95% CI, −10.5 to 11.8) or high-risk disease (absolute difference, 2.3 percentage points; 95% CI, −11.5 to 16.1) (P=0.08 for interaction). Treatment for disease progression was less frequent with surgery than with observation (absolute difference, 26.2 percentage points; 95% CI, 19.0 to 32.9); treatment was primarily for asymptomatic, local, or biochemical (prostate- specific antigen) progression. Urinary incontinence and erectile and sexual dysfunction were each greater with surgery than with observation through 10 years. Disease-related or treatment-related limitations in activities of daily living were greater with surgery than with observation through 2 years.

CONCLUSIONS

After nearly 20 years of follow-up among men with localized prostate cancer, surgery was not associated with significantly lower all-cause or prostate-cancer mortality than observation. Surgery was associated with a higher frequency of adverse events than observation but a lower frequency of treatment for disease progression, mostly for asymptomatic, local, or biochemical progression. (Funded by the Department of Veter- ans Affairs and others; PIVOT  ClinicalTrials.gov number, NCT00007644.)

Prostatectomy vs. Observation for Prostate Cancer

We previously reported that radical prostatectomy was not associated with significantly lower all-cause or prostate-cancer mortality than observation with noncurative palliative interventions through 12 years among men with localized prostate cancer detected during the early era of prostate-specific antigen (PSA) testing.1 In light of the protracted disease course and extended survival of many patients, treatment decisions often require infor- mation about additional treatments received, patient-reported outcomes, and very-long-term mortality. Three other randomized trials com- pared radical prostatectomy with observation or PSA-based active monitoring. One older trial showed no significant difference in overall mor- tality.2 Another trial, also conducted before the widespread use of PSA screening, showed differ- ences favoring surgery in all-cause and prostate- cancer mortality of 12.7 and 11.0 percentage points, respectively, with similar differences in the risk of distant metastases at a median follow- up of 13.4 years.3 The most recent trial, involving men with disease detected by PSA screening, showed no significant difference in all-cause or prostate-cancer mortality after a median of 10 years of follow-up among radiation therapy, sur- gery, and PSA-based active monitoring and de- layed radical intervention.4 We report all-cause and prostate-cancer mortality through nearly 20 years of follow-up and describe disease progres- sion, treatments received, and patient-reported outcomes during the original follow-up.

Trial Design

We previously described the design, methods, and baseline results of the Prostate Cancer Interven- tion versus Observation Trial (PIVOT).1,5 The trial was approved by the institutional review board at each site. Patients provided written informed consent. After completion of follow-up through January 2010, we amended the protocol to assess extended all-cause and prostate-cancer mortality. The original and revised protocols, including the statistical analysis plan, are available with the full text of this article at NEJM.org.

Patients

From November 1994 through January 2002, we randomly assigned 731 men with localized pros- tate cancer (mean age, 67 years; median PSA value, 7.8 ng per milliliter) to radical prostatec- tomy or observation at Department of Veterans Affairs and National Cancer Institute medical centers.1,5 Patients had to be medically fit for radical prostatectomy and have histologically confirmed, clinically localized prostate cancer (stage T1-T2NxM0 in the tumor–node–metasta- sis classification system according to the Ameri- can Joint Committee on Cancer6) of any grade diagnosed within the previous 12 months. Pa- tients had to have a PSA value of less than 50 ng per milliliter, an age of 75 years or younger, negative results on a bone scan for metastatic disease, and a life expectancy of at least 10 years. Trial sites assessed eligibility on the basis of lo- cal PSA values and biopsy readings. One of the authors, who is a pathologist, reviewed and re- classified biopsy specimens at a central location. The pathologist was unaware of the long-term fate of the patient and of the original Gleason score assessment at the local center.

Follow-up and Clinical Outcomes

We assessed all-cause and prostate-cancer mor- tality through August 2014, for a minimum of 12 years and a maximum of 19.5 years or until the patient died. Logistic and analytic difficul- ties, including difficulty in obtaining informa- tion about and adjudicating the cause of death, delayed the reporting of findings. An end-points committee whose members were unaware of the trial-group assignments determined the cause of death on the basis of information extracted from the patients’ medical records.1,7 Data on disease progression, treatment received, and patient- reported health outcomes were based on follow- up through January 2010. After January 2010, data on treatment and follow-up were not system- atically collected. We classified progression as local, regional, systemic, or biochemical (PSA) and as symptomatic or asymptomatic (Table S1 in the Supplementary Appendix, available at NEJM.org). We defined treatments as “definitive intervention” (Fig. S1 in the Supplementary Appen- dix), any surgery (including radical prostatectomy), radiation therapy, hormone therapy, chemother- apy, or immunotherapy, and we report treat- ments according to type of disease progression.

Statistical Analysis

Our primary outcome was all-cause mortality. Our secondary outcome was prostate-cancer mortality, which was defined as death that was definitely

n engl j med 377;2    nejm.org    July 13, 2017                                                                                                                                  133

T he New England Journal of Medicine

peting risks, because the Kaplan–Meier method for unadjusted survival analysis yields unreliable results for survival estimates in the presence of competing risks.8-10 We assessed the cumulative incidence of death, between-group differences, and relative risks at 4, 8, 12, and 16 years and at the end of follow-up. P values of less than 0.05 were considered to indicate statistical signifi- cance, with no adjustment for multiple compari- sons. We prespecified seven subgroups accord- ing to baseline characteristics, as described previously1: age, race, coexisting conditions,11 self- reported performance status, PSA level, Gleason score,12 and D’Amico tumor risk score.13 We also assessed mortality outcomes on the basis of central histopathological reclassification. Patient- reported outcomes were analyzed at baseline, 6 months, 1 year, 2 years, 5 years, and 10 years with the use of Fisher’s exact test for categorical outcomes or a t-test for continuous measures. We used SAS software, versions 9.3 and 9.4 (SAS Institute), for the  analyses.14 or probably due to prostate cancer or definitely or probably due to treatment for prostate cancer. End points through January 2010 included the following: local, regional, systemic, and bio- chemical (PSA) progression; additional treat- ments; adverse events requiring treatment; and patient-reported outcomes of urinary inconti- nence, erectile and sexual dysfunction, worry about health, “bother” due to prostate cancer or treatment, physical discomfort, satisfaction with sexual functioning, and functional limitations due to prostate cancer or treatment.

Analyses were performed according to the intention-to-treat principle with the use of Kaplan– Meier methods and corresponding 95% confi- dence intervals. We used the Fine and Gray method to compare groups with respect to prostate-cancer mortality in the presence of  com-

All-Cause and Prostate-Cancer Mortality

As of August 2014, a total of 468 of 731 men (64.0%) had died. The vital status of all the par- ticipants was available, although we were unable to ascertain the cause of death in 7 men (2 as- signed to surgery and 5 to observation). The median follow-up from randomization to death or the end of follow-up was 12.7 years (inter- quartile range, 7.3 to 15.5). All-cause mortality was not significantly lower with surgery than with observation (hazard ratio, 0.84; 95% confi- dence interval [CI], 0.70 to 1.01; P=0.06) (Fig. 1A). The 19.5-year cumulative incidence of death was 61.3% among men assigned to radical prostatec- tomy and 66.8% among those assigned to obser- vation (relative risk, 0.92; 95% CI, 0.82 to 1.02) (Table 1). Absolute differences in risk, although not significant, increased from 3.1 percentage points at 8 years to 5.5 percentage points at the end of follow-up (Table S3 in the Supplementary Appendix). The median survival was 13.0 years (95% CI, 12.5 to 13.5) with surgery and 12.4 years

(95% CI, 11.4 to 12.8) with observation.

Death attributed to prostate cancer or treat- ment occurred in 69 men (9.4%); 65 deaths were attributed to prostate cancer and 4 to treatment. Prostate-cancer mortality was not significantly

134                                                                                 n engl j med 377;2    nejm.org    July 13, 2017

Prostatectomy vs. Observation for Prostate Cancer

* PSA  denotes  prostate-specific antigen.

† Race was reported by the participants.

‡ The risk category was determined according to the D’Amico risk score, which is based on tumor stage, histologic score, and PSA level.  lower with radical prostatectomy than with ob- servation (hazard ratio, 0.63; 95% CI, 0.39 to 1.02; P= 0.06) (Fig. 1B). The cumulative incidence of death due to prostate cancer or treatment was 7.4% with surgery and 11.4% with observation (absolute  difference, 4.0  percentage points; 95% CI, −0.2 to 8.3) (Table 2). Relative and absolute differences in risk remained stable after 12 years (Table S4 in the Supplementary Appendix). Death that was considered to be definitely due to prostate cancer or treatment occurred in 18 men (4.9%) assigned to surgery and 22 men (6.0%) assigned to observation.

Subgroup Analyses

The effect of surgery on mortality did not differ significantly according to baseline patient characteristics (Fig. 2A and 2B and Tables 1 and 2, and Tables S3 and S4 in the Supplementary Ap- pendix). Among men younger than 65 years of age, the absolute difference in all-cause mortal- ity between the surgery group and the observa- tion group increased from −1.5 percentage points (95% CI, −11.7 to 8.5) at 8 years to 12.0 percent- age points (95% CI, −0.3 to 23.8) at the end of follow-up. Among men 65 years of age or older, the absolute difference at the end of follow-up was 2.6 percentage points (95% CI, −5.7 to 10.8). The effect of surgery on all-cause mortality may have differed according to baseline PSA value (P = 0.06 for interaction) and tumor risk category (P= 0.08 for interaction) (Fig. 2A and Table 1, and Table S3 in the Supplementary Ap- pendix). Surgery was not associated with lower

n engl j med 377;2    nejm.org    July 13, 2017                                                                                                                                  135

T he New England Journal of Medicine

all-cause mortality than observation among men with a PSA value of 10 ng per milliliter or less or among those with low-risk or high-risk cancers but may have been associated with lower mortal- ity among men with a PSA value of more than 10 ng per milliliter or among those with intermedi- ate-risk disease. Among men with low-risk dis- ease, the absolute difference in risk between the trial groups was 0.7 percentage points (95% CI, −10.5 to 11.8). Among men with intermediate- risk tumors, the absolute difference in risk was 14.5 percentage points (95% CI, 2.8 to 25.6). The absolute difference in risk was smaller and not significant among men with disease that was determined to be intermediate-risk on the basis of central (rather than local) Gleason  grading.

136                                                                                 n engl j med 377;2    nejm.org    July 13, 2017

Prostatectomy vs. Observation for Prostate Cancer

n engl j med 377;2    nejm.org    July 13, 2017                                                                                                                                  137

T he New England Journal of Medicine

* Disease progression was defined according to the trial protocol as the time to evidence of disease progression or persistence. Progression  was classified according to clinical stage (local, regional, or systemic) and whether asymptomatic or causing clinical signs or symptoms.

† Data reflect recorded treatments and indications for treatment on the basis of clinical stage, including an increasing or persistently elevated PSA level. A patient could be counted as having received treatment for more than a single type of disease progression.

‡ Data are based on treatments with a known start date.

Exploratory analyses in men with T1c (nonpal- pable, PSA-detected) tumors, including those with a PSA value of more than 10 ng per milliliter or a Gleason score of 7 or more (on a scale from 2 to 10, with 10 indicating the most poorly differen- tiated tumors), showed nonsignificant differ- ences in all-cause and prostate-cancer mortality between the surgery group and the observation group (Tables S3 and S4 in the Supplementary Appendix). Few deaths from prostate cancer or treatment occurred among men with T1c tumors, and confidence intervals were wide (Tables S3 and S4 in the Supplementary Appendix).

Disease Progression and Additional Treatments

Fewer men assigned to surgery had disease pro- gression or received additional treatment than men assigned to observation (Table 3, and Figs. S2 through S4 in the Supplementary Appendix). Any progression occurred in 40.9% of the men assigned to surgery versus 68.4% of the men as- signed to observation. Most disease progression was local, and approximately half the cases of local progression were asymptomatic. Definitive treatment occurred in 20.4% of the men assigned to observation, rarely after 5 years of follow-up,

138                                                                                 n engl j med 377;2    nejm.org    July 13, 2017

Prostatectomy vs. Observation for Prostate Cancer

and in 85.5% of the men assigned to surgery, almost all occurring within 1 year (Fig. S1 in the Supplementary Appendix). Treatment for cancer progression, mostly for asymptomatic, local, or PSA progression, occurred in 33.5% of the men assigned to surgery and in 59.7% of the men as- signed to observation (Table 3, and Fig. S3 in the Supplementary Appendix). Androgen-deprivation therapy was less frequent among men assigned to surgery than among those assigned to obser- vation (21.7% vs. 44.4%). The absolute difference in the risk of systemic progression or treatment for systemic progression was approximately 4 per- centage points, in favor of surgery.

The frequency of treatment for local progres- sion was lower with surgery than with observa- tion in all tumor risk groups. Among men with low-risk disease, treatment for regional or sys- temic disease was infrequent and did not differ significantly between the surgery group and the observation group. Among men with intermedi- ate-risk disease, the frequency of treatment for systemic progression was lower with surgery than with observation (5.4% vs. 11.7%). Among men with high-risk disease, the frequency of treat- ment for an increasing or persistently elevated PSA value was lower with surgery than with observation (33.8% vs. 52.5%), but there was no significant difference with respect to treatment for regional or systemic progression. (Details re- garding treatment according to progression type and tumor risk category are provided in Fig. S4 in the Supplementary Appendix.) Physician-pre- scribed treatment for erectile dysfunction and urinary incontinence due to prostate-cancer pro- gression or treatment was more common with surgery than with observation (Table  3).

Patient-Reported Outcomes

Patient-reported overall health, physical or men- tal health assessed with the use of the Medical Outcomes Study 12-Item Short-Form General Health Survey (SF-12), and worry about health did not differ significantly between the groups. As compared with men assigned to observation, men assigned to surgery were more likely to re- port bother due to prostate cancer or treatment, physical discomfort, and limitations in activities of daily living through 2 years but not at later time points. The use of absorbent pads because of incontinence was greater through 10 years in men treated with surgery than in those assigned

to observation; absolute differences exceeded 30 percentage points at all time points. Erectile dys- function as well as decreases in sexual function, activity, interest, and satisfaction were much greater through 5 years in men assigned to sur- gery than in those assigned to observation. De- tails on patient-related outcomes are provided in Figures S5 through S9 and Tables S5 through S7 in the Supplementary Appendix.

After nearly 20 years, the absolute difference in all-cause mortality between men assigned to sur- gery and those assigned to observation was less than 6 percentage points, and the absolute dif- ference in prostate-cancer mortality was 4 per- centage points. The extended follow-up yielded slightly greater differences in mortality favoring surgery than those described earlier, although the differences remained nonsignificant.1,14-16 The frequencies of disease progression and treatment for disease progression were lower with surgery than with observation, although most events of progression were asymptomatic, local, or bio- chemical. There were large, long-term differ- ences in urinary incontinence and erectile and sexual dysfunction in favor of observation and smaller, shorter-term differences in adverse ef- fects with respect to physical function and activi- ties of daily living.

Although relative differences in prostate-cancer mortality between the trial groups appeared to be larger, absolute differences were small, both relative and absolute differences were nonsignifi- cant, and statistical measures were not adjusted for multiple comparisons; such an adjustment would have further reduced the levels of signifi- cance of the differences.17 We encourage readers to focus on absolute differences in risk and cor- responding confidence intervals.14-16 The differ- ence was 1 percentage point when the end point was death that was considered to be definitely due to prostate cancer or treatment. Using all- cause mortality as our primary end point avoid- ed pitfalls in cause-of-death  ascertainment.7,18

We urge caution in interpreting subgroup effects.19 Our trial was not powered to detect small differences between subgroups, but significant results may be due to multiple analyses.17 If differences in all-cause mortality exist, they were confined to men with intermediate-risk disease

n engl j med 377;2    nejm.org    July 13, 2017                                                                                                                                  139

T he New England Journal of Medicine

and depend on histopathological classification methods. For men with low-risk or high-risk disease, differences in all-cause mortality were 3 percentage points or less and were not signifi- cant. Surgical effects on mortality did not vary according to patient factors.

Most treatment for disease progression was for local, asymptomatic reasons, especially in- creasing or persistently elevated PSA values. Among men assigned to surgery, 41% had dis- ease progression and 34% received treatment for disease progression. There were few differences in quality of life and worry about health between the surgery group and the observation group. Long-term erectile and sexual dysfunction and urinary incontinence and physician-directed treat- ment were substantially greater with surgery than with observation. Functional limitations and bother due to prostate cancer or treatment were greater with surgery than with observation through at least 2 years, whereas worry about health was less at 10 years.

Our findings are generally consistent with those of the Scandinavian Prostate Cancer Group Study Number 4 (SPCG-4)3,20 and the Prostate Testing for Cancer and Treatment (ProtecT) trial.4,21 SPCG-4, like PIVOT, compared surgery with ob- servation, whereas the ProtecT trial compared surgery or radiation with active monitoring and delayed radical intervention based primarily on PSA results without surveillance biopsies. Data from randomized trials are lacking to assess biopsy-based active surveillance, the predominant active-monitoring practice in the United States. PIVOT involved men who received a diagnosis of prostate cancer, and who used treatments avail- able, in the United States during the early era of PSA testing,1,5 a midpoint between the era before PSA testing (SPCG-4) and the later era of PSA testing (the ProtecT trial). Mortality differences across these studies may reflect differences in patient age and coexisting conditions but more likely reflect differences in the natural history of prostate cancer that are related to detection methods and possibly treatment approaches, including improvements in medical treatment for progressive disease. Only 9.4% of PIVOT participants died from prostate cancer. The ab- solute difference in all-cause mortality between the trial groups was 5.5 percentage points in favor of surgery, and the absolute difference in prostate-cancer mortality was 4.0  percentage points. In the ProtecT trial, mortality and the incidence of metastatic disease (which included asymptomatic disease detected by surveillance imaging and biochemical [PSA] testing) as well as absolute differences between the trial groups in favor of radical intervention (1 percentage point for mortality and 4 percentage points for metastatic-disease incidence) were much lower through 10 years than in SPCG-4 or PIVOT.

Our results, together with those of SPCG-4, the ProtecT trial, and two earlier trials that showed no mortality benefits of surgery2 or radia- tion22 as compared with observation, have clini- cal implications. First, they show that long-term prostate-cancer mortality remains low among most men with localized prostate cancer who are treated with observation and that death from prostate cancer is very uncommon among men with low-risk and low-PSA disease. Reducing over- treatment is needed. Men with low-risk and PSA- based screening–detected disease can safely avoid harms and costs of early radical intervention or of biopsy-guided active surveillance with delayed radical treatment.4,21,23 Observation, PSA-based monitoring, and active surveillance with delayed radical intervention remain infrequently used, even among older men,24-26 despite a frequency of metastatic progression of less than 3%,27 prostate- cancer mortality of 1% or less,28-31 and cost-effec- tiveness that is superior to that with early radical intervention.23,32 PSA-based monitoring and biopsy- based active-surveillance programs should reduce the frequency of surveillance biopsy and increase biopsy and PSA thresholds that trigger radical interventions.

Second, surgery may have mortality benefits in some men, particularly those with intermediate- risk prostate cancer who have long life expectan- cies. However, the comparative effects of active surveillance and PSA-based monitoring in many men with intermediate-risk disease should be examined. In addition, the risk of progression or death as well as the absolute treatment benefit diminish while overtreatment harms increase in men with smaller-volume, screening-detected lower-risk cancers. Beneficial effects depend on proper histopathological grading, which is fraught with interobserver and intraobserver variation. Revisions in Gleason grading and risk-score clas- sification systems have resulted in an upgrading of prostate cancers that are classified today, as compared with those in men enrolled in PIVOT.

140                                                                                 n engl j med 377;2    nejm.org    July 13, 2017

Prostatectomy vs. Observation for Prostate Cancer

This has led to more men being classified with intermediate-risk or high-risk disease, resulting in fewer men being considered eligible for ob- servation or active  monitoring.33-36

Third, although men with high-risk disease have a poor prognosis, surgery may not provide large benefits with respect to mortality. Safer and more effective options are needed. Fourth, surgery is associated with a decreased risk of disease progression and treatment for disease progression. However, most progression is asymp- tomatic, local, or biochemical, for which the treatment benefit is uncertain. Among men with low-risk disease in our trial, the frequency of treatment for regional or systemic progression was not significantly lower with surgery than with observation. Reducing treatment for asymp- tomatic progression would decrease harms and costs, with little, if any, effect on mortality. Fifth, surgery causes perioperative and longer- term adverse effects, some requiring treatment. Nonetheless, regardless of the initial treatment, we found few differences between the trial groups in long-term bother, physical discomfort, worry about health, overall health, or limitations in activities due to prostate cancer or treatment.

Differences in satisfaction with sexual function- ing remained significant through 5 years.

In conclusion, radical prostatectomy was not associated with significantly lower all-cause or prostate-cancer mortality than observation through 20 years of follow-up among men with localized prostate cancer that was diagnosed during the early era of PSA testing. Absolute differences re- mained below 6 percentage points. Death from prostate cancer was very uncommon among men with low-risk disease who were assigned to observation. Surgery may be associated with de- creased mortality among men with intermediate- risk prostate cancer, depending on the pathologi- cal classification. Surgery resulted in substantially greater long-term urinary incontinence and erec- tile and sexual dysfunction than observation and was associated with a significantly lower risk of disease progression and additional treatments, most for local or asymptomatic biochemical progression.

Supported by the Department of Veterans Affairs, the Agency for Healthcare Quality and Research, and the National Cancer Institute.

Disclosure forms provided by the authors are available with the full text of this article at NEJM.org

References

1. Wilt TJ, Brawer MK, Jones KM, et al. Radical prostatectomy versus observation for localized prostate N Engl J Med 2012;367:203-13.
2. Iversen P, Madsen PO, Corle DK. Rad- ical prostatectomy versus expectant treat- ment for early carcinoma of the prostate: twenty-three year follow-up of a prospec- tive randomized study. Scand J Urol Nephrol Suppl 1995;172:65-72.
3. Bill-Axelson A, Holmberg L, Garmo H, et al. Radical prostatectomy or watch- ful waiting in early prostate N Engl J Med 2014;370:932-42.
4. Hamdy FC, Donovan JL, Lane JA, et 10-Year outcomes after monitoring, sur- gery, or radiotherapy for localized pros- tate cancer. N Engl J Med 2016;375:1415- 24.
5. Wilt TJ, Brawer MK, Barry MJ, et al. The Prostate cancer Intervention Versus Observation Trial:VA/NCI/AHRQ Cooper- ative Studies Program #407 (PIVOT): de- sign and baseline results of a randomized controlled trial comparing radical prosta- tectomy to watchful waiting for men with clinically localized prostate cancer. Con- temp Clin Trials 2009;30:81-7.
6. Fleming ID, Cooper JS, Henson DE, et al., eds. AJCC cancer staging manual. 5th Philadelphia: Lippincott-Raven, 1997.
7. Barry MJ, Andriole GL, Culkin DJ, et

Ascertaining cause of death among men in the Prostate Cancer Intervention versus Observation Trial. Clin Trials 2013;10: 907-14.

8. Lin G, So Y, Johnston G. Analyzing survival data with competing risks using SAS Cary, NC: SAS Institute, 2012 (https://support.sas.com/resources/ papers/proceedings12/344-2012.pdf).
9. Gray RJ. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat 1988;16:1141-
10. Albertsen PC, Hanley JA, Gleason DF, Barry Competing risk analysis of men aged 55 to 74 years at diagnosis managed conservatively for clinically localized pros- tate cancer. JAMA 1998;280:975-80
11. Charlson ME, Pompei P, Ales KL, MacKenzie A new method of classify- ing prognostic comorbidity in longitudi- nal studies: development and validation. J Chronic Dis 1987;40:373-83.
12. Gleason DF, Veteran’s Administration Cooperative Urologic Research Group. Histologic grading and clinical staging of prostatic carcinoma. In: Tannenbaum M, ed. Urologic pathology: the Phil- adelphia: Lea & Febiger, 1977:171-98.
13. D’Amico AV, Whittington R, Malko- wicz SB, et Biochemical outcome after radical prostatectomy, external beam ra-

diation therapy, or interstitial radiation therapy for clinically localized prostate cancer.  JAMA 1998;280:969-74.

14. Allison PD. Survival analysis using the SAS system: a practical guide. Cary, NC: SAS Institute, 1995:292.
15. Pocock SJ, Stone The primary outcome fails — what next? N Engl J Med 2016;375:861-70.
16. Pocock SJ, Stone The primary out- come is positive — is that good enough? N Engl J Med 2016;375:971-9.
17. Chavalarias D, Wallach JD, Li AHT, Ioannidis Evolution of reporting P val- ues in the biomedical literature, 1990-2015. JAMA 2016;315:1141-8.
18. Newschaffer CJ, Otani K, McDonald MK, Penberthy Causes of death in el- derly prostate cancer patients and in a comparison nonprostate cancer cohort. J Natl Cancer Inst 2000;92:613-21.
19. Wang R, Lagakos SW, Ware JH, Hunter DJ, Drazen JM. Statistics in medicine — reporting of subgroup analyses in clinical trials. N Engl J Med 2007;357:2189-94.
20. Johansson E, Steineck G, Holmberg L, et al. Long-term quality-of-life outcomes after radical prostatectomy or watchful waiting: the Scandinavian Prostate Cancer Group-4 randomised trial. Lancet Oncol 2011;12:891-9.
21. Donovan JL, Hamdy FC, Lane JA, et

n engl j med 377;2    nejm.org    July 13, 2017                                                                                                                                  141

Prostatectomy vs. Observation for Prostate Cancer

Patient-reported outcomes after monitor- ing, surgery, or radiotherapy for prostate cancer. N Engl J Med 2016;375:1425-37.

22. Sanyal C, Aprikian AG, Cury FL, Che- valier S, Dragomir A. Management of localized and advanced prostate cancer in Canada: a lifetime cost and quality- adjusted life-year analysis. Cancer 2016; 122:1085-96.
23. Maurice MJ, Zhu H,  Abouassaly    A hospital-based study of initial observa- tion for low-risk prostate cancer and its predictors in the United States. Can Urol Assoc J 2015;9:E193-E199.
24. Maurice MJ, Abouassaly R, Kim SP, Zhu H, Hui Zhu H. Contemporary nation- wide patterns of active surveillance use for prostate cancer. JAMA Intern Med 2015;175:1569-71.
25. Hoffman KE, Niu J, Shen Y, et al. Phy- sician variation in management of low- risk prostate cancer: a population-based cohort JAMA Intern Med 2014;174: 1450-9.
26. Yamamoto T, Musunuru B,  Vesprini

D, et al. Metastatic prostate cancer in men initially treated with active surveillance. J Urol 2016;195:1409-14.

27. Dall’Era MA, Albertsen PC, Bangma C, et Active surveillance for prostate can- cer: a systematic review of the literature. Eur Urol 2012;62:976-83.
28. Godtman RA, Holmberg E, Khatami A, Stranne J, Hugosson Outcome follow- ing active surveillance of men with screen- detected prostate cancer: results from the Göteborg randomised population-based prostate cancer screening trial. Eur Urol 2013;63:101-7.
29. Klotz L, Vesprini D, Sethukavalan P, et Long-term follow-up of a large active surveillance cohort of patients with pros- tate cancer. J Clin Oncol 2015;33:272-7.
30. Tosoian JJ, Mamawala M, Epstein JI, et al. Intermediate and longer-term out- comes from a prospective active-surveil- lance program for favorable-risk prostate J Clin Oncol 2015;33:3379-85.
31. Hayes JH, Ollendorf DA, Pearson SD, et Observation versus initial treatment

for men with localized, low-risk prostate cancer: a cost-effectiveness analysis. Ann Intern Med 2013;158:853-60.

32. Helpap B, Egevad L. The significance of modified Gleason grading of prostatic carcinoma in biopsy and radical prosta- tectomy Virchows Arch 2006; 449:622-7.
33. Epstein JI. An update of the Gleason grading system. J Urol 2010;183:433-40.
34. Albertsen PC, Hanley JA, Barrows GH, et al. Prostate cancer and the Will Rogers phenomenon. J Natl Cancer Inst 2005;97:1248-53.
35. Cooperberg MR, Carroll Trends in management for patients with localized prostate cancer, 1990-2013. JAMA 2015; 314:80-2.
36. S. Preventive Services Task Force. Prostate cancer: screening — draft rec- ommendation statement (https://www

.uspreventiveservicestaskforce.org/Page/ Document/draft-recommendation

-statement/prostate-cancer-screening1).

Copyright © 2017 Massachusetts Medical Society.