Full publication of the results of the PIVOT study


The complete results of the Prostate Cancer Intervention versus Observation Trial (PIVOT) study have (finally) just been published in the New England Journal of Medicine. Unfortunately, only the abstract is available on line so it may take us a little while to get hold of a copy of the complete report.

For those who may have forgotten (or simply not been aware of this study), this was a long-term study of the relative benefits (and risks) of radical prostatectomy to simple “observation” (effectively a watchful waiting strategy as opposed to an active surveillance strategy) in men initially diagnosed with localized prostate cancer.

According to the abstract (and as previously reported):

  • The trial enrolled 731 men with localized prostate cancer between November 1994 and January 2002.
  • These 731 men were randomly assigned to either be treated with a radical prostatectomy (RP) or were simply “observed.”
  • All patients were followed until January 2010.
  • The primary study outcome was all-cause mortality.
  • The secondary study outcome was prostate cancer-specific mortality.

The core results of the study are as follows:

  • Average (mean) age of the patients at enrollment was 67 years.
  • Average (median) PSA level of the patients at enrollment was 7.8 ng/ml.
  • Average (median) follow-up was 10.0 years.
  • During the follow-up period
    • 171/364 patients (47.0 percent) randomized to RP died.
    • 183/367 patients (49.9 percent) randomized to observation also died.
    • The absolute reduction in risk of death as a consequence of RP as opposed to observation was 2.9 percent (hazard ratio [HR] = 0.88).
    • 21/364 patients (5.8 percent) randomized to RP died either from their prostate cancer or as a direct consequence of their treatment.
    • 31/367 patients (8.4 percent) randomized to observation died from prostate cancer.
    • The absolute reduction in risk of death as a consequence of RP as opposed to observation was 2.6 percent (HR = 0.63).
  • Adverse events after surgery occurred within 30 days in 21.4 percent of patients treated by RP, inclusive of one death.
  • Among men with a baseline PSA > 10 ng/ml, RP was associated with a statistically significant  reduction in risk for all-cause mortality.
  • Among men with intermediate-risk or high-risk prostate cancer there appears to have been the possibility of a statistically significant reduction in risk for all-cause mortality.
  • There was no impact of treatment on all-cause and prostate-cancer mortality according to age, race, coexisting conditions, self-reported performance status, or histologic features of the tumor.

The authors conclude with the careful statement that:

Among men with localized prostate cancer detected during the early era of PSA testing, radical prostatectomy did not significantly reduce all-cause or prostate-cancer mortality, as compared with observation, through at least 12 years of follow-up.

There appear to be no important differences in the final abstract of this paper as compared to data originally reported by Wilt at the annual meeting of the American Urological Association in May 2011, which only makes it more surprising that it has taken so long for this report to be published.

We will reserve further comment on this paper until we have had the chance to read the entire text of the article, as well as the entire text of the associated editorial by Thompson and Tangen entitled “Prostate cancer — uncertainty and a way forward.” More information is available in a commentary already available on the MedPage Today web site. (They clearly had early access to the full text of the article and the associated editorial.)

10 Responses

  1. I’ve read the full study and the appendix, and have a few comments.

    1. First, it is certainly the case that the study shows weaker effects on mortality in the full sample than some might have expected. At 12 years after diagnosis, we have prostate cancer mortality in the full sample dropping from 7.4% to 4.4%, and all-cause mortality dropping from 44% to 41%. Given the modest sample size, these drops in mortality are not statistically significantly different from zero. On the other hand, given the small sample size, we also can’t statistically reject that the prostate cancer morality reductions might be as great as 6%, and the all-cause mortality reductions as great as 10%. Sample size is a problem in the study, which reflects in part that they did not end up getting a sample size as large as they had originally hoped for.

    2. Second, the results, strictly speaking, only apply to groups that resemble the study group. The study group appears to be somewhat unusual in one respect, and very unusual in a second respect. The “somewhat unusual” aspect is that this is an older sample — 90% 60 or older; mean age of 67. So it is unclear to what extent these results would apply to a 55-year-old man. The more unusual aspect of the sample is that the study was only able to get 15% participation. They approached 5,023 men diagnosed with prostate cancer, and only 731 agreed to be randomized to surgery vs. observation, and the other 4,292 refused to be part of the study. So the study sample is highly selective. One has to ask what might be different among men who chose to participate in this study versus those who did not. The study text and appendix has no discussion of possible selection bias due to this selection, and does not discuss what observable variables are correlated with why men participated in this study. One might speculate that, for example, men who had a relative die of prostate cancer might be less likely to agree to be randomized.

    On the whole, I think the results indicate that among men in their 60s or older, they should lean against surgery and towards active surveillance if they are in a low-risk group, which might include whether they had no relatives who died from prostate cancer. On the other hand, men in medium-risk and high-risk groups seem to have greater mortality reductions due to prostate cancer, and so perhaps the lean there goes in the other direction. If this advice were followed, it would probably reduce the number of surgeries in this older age group. I don’t think the results have strong implications one way or the other for men in their 50s or men who have relatives who died of prostate cancer, as I suspect these groups are not well represented in the study sample.

  2. Thanks Tim.

  3. This study differs only slightly from the Scandinavian study that showed a 5.8% mortality reduction at 14 years, and in that study there was no benefit for men over 65. At the very least, these results should be presented to all men diagnosed with the disease. The problem will remain, however, that despite these results, few men faced with a diagnosis of even low-risk cancer will be comfortable treating it conservatively. This means that all the results should be presented before a man even takes a PSA test so they have a chance to decide whether to get tested and take on the anxiety associated with the whole process.

    As for the previous comments, the challenge of what to tell younger men will remain. Should they assume that being younger they are more likely to benefit or that, without proof of benefit, they should be more conservative to avoid the negative impact on their quality of life?

  4. I just want to point out what I think is an error in the list of “core results” from the study: The fifth open bullet says, “31/367 patients (8.4 percent) randomized to RP died from prostate cancer.” Shouldn’t this say “31/367 patients (8.4 percent) randomized to OBSERVATION died from prostate cancer” instead?

  5. Thank you Doug. You are quite correct. I clearly wasn’t fully awake at 7 am this morning! This will be corrected in seconds.

  6. It still remains, that nobody can yet appropriately advise a patient in his mid-50s who has prostate cancer that is not aggressive, and asks the simple question: “If we just watch it, how will we know when it is time to act? How will we know when it becomes too late to act? When does my window of opportunity close?”

    For this reason, I chose RALP and I am not sorry I did. I simply had too many healthy years ahead of me to play a waiting game, always wondering when that window would slam shut on me, while the doctors and I just watched and waited.

  7. Very nice summary of the study and appreciate the comments by Tim Bartik. As a 50-year-old diagnosed with medium-risk prostate cancer (Gleason 7), this study (and the USPSTF recommendation not to screen for prostate cancer) don’t help me know what to do. It still comes down to a risk/benefit decision for me, and I prefer to make that choice now rather than wait until I am 10 or 15 years older and possibly have fewer choices.

  8. I cannot see a great significance from this trial’s results for currently diagnosed men to make a treatment/no treatment decision. The fact that they could not get enough participants to provide the necessary statistical power cannot be ignored. Patients were randomized at 44 Veterans Affairs sites and eight National Cancer Institute sites.

    Primary entry criteria were: age < 75; clinically localized prostate cancer (T1-T2NxM0); PSA < 50 ng/mL; negative bone scan; life expectancy of at least 10 years. Among men with complete data, 40% had low-risk cancers, 34% intermediate risk, and 21% high risk

    In spite of this, overall mortality was 6.6% lower in the surgical series and disease-specific mortality was 31% lower in the surgical cohort. Maybe not statistically significant, but still an eye-opener given the results expressed by the authors. They tell a different story even when they took their time to do it.

  9. “As a 50-year-old diagnosed with medium-risk prostate cancer (Gleason 7), this study doesn’t help me know what to do.”

    Every 50-year-old (diagnosis or not) should have the option of a preventive prostatectomy. There is the potential (in Australia) of saving one life for every 26 prostatectomies.

  10. “There is the potential (in Australia) of saving one life for every 26 prostatectomies.”

    That should have been saving one life for every 23 prostatectomies. I was relying on memory.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s