Correspondence

Prostatectomy or Watchful Waiting in Prostate Cancer

N Engl J Med 2003; 348:170-171January 9, 2003

Article

To the Editor:

In their article comparing prostatectomy with watchful waiting, Holmberg et al. (Sept. 12 issue)1 present their analysis according to the intention-to-treat principle, but the true treatment effect might be larger than that observed because of noncompliance with the protocol.2 Presentation of information on the outcome in the participants who did not comply with the protocol and an additional per-protocol analysis could give more insight into the real treatment effect.

In addition, the authors report a statistically significant difference between the two groups in terms of distant metastases. However, outside of the protocol, untreated men might undergo a search for metastases more often than treated men. Such a difference could have serious implications for the conclusions drawn by the authors.

Marco H. Blanker, M.D., Ph.D.
Sita M.A. Bierma-Zeinstra, Ph.D.
Boris W.V. Schouten, M.D.
University Medical Center Rotterdam, 3000 DR Rotterdam, the Netherlands
blanker@hag.fgg.eur.nl

2 References

1. 1

   Holmberg L, Bill-Axelson A, Helgesen F, et al. A randomized trial comparing radical prostatectomy with watchful waiting in early prostate cancer. N Engl J Med 2002;347:781-789
   Full Text | Web of Science | Medline

2. 2

   Nagelkerke N, Fidler V, Bernsen R, Borgdorff M. Estimating treatment effects in randomized clinical trials in the presence of non-compliance. Stat Med 2000;19:1849-1864[Erratum, Stat Med 2001;20:982.]
   CrossRef | Web of Science | Medline

To the Editor:

Holmberg et al. showed that radical prostatectomy reduces the likelihood of distant metastases and the need for androgen-deprivation therapy, as compared with watchful waiting. Androgen-deprivation therapy is associated with morbidity in men with prostate cancer and represents an economic burden on both patients and the health care system. Thus, freedom from castration represents an important end point after local therapy.1 Local therapy for prostate cancer is effective in decreasing the rate of distant metastasis and averting the need for androgen-deprivation therapy. These end points alone may be sufficient justification for local therapy, even in the absence of a gain in overall survival.

John J. Coen, M.D.
Anthony L. Zietman, M.D.
William U. Shipley, M.D.
Massachusetts General Hospital, Boston, MA 02114
jcoen@partners.org

1 References

1. 1

   Zietman A, Thakral H, Skowronski U, Shipley W. Freedom from castration: an alternative end point for men with localized prostate cancer treated by external beam radiation therapy. Int J Radiat Oncol Biol Phys 2002;53:1152-1159
   CrossRef | Web of Science | Medline

To the Editor:

Holmberg and colleagues do not report data for patients according to the Gleason score. Are the conclusions of their study also valid for patients with aggressive cancers, such as those with a Gleason score of more than 7?

Carlo B. Gambacorti-Passerini, M.D.
National Cancer Institute, 20133 Milan, Italy
gambacorti@istitutotumori.mi.it

Elisabetta A. Sala, M.D.
Via Oslavia 25, 20052 Monza, Italy

To the Editor:

I believe the primary end point of death due to prostate cancer was not reached by a significant number of patients in either group in the study by Holmberg et al. Less than 10 percent of patients in each group died from progressive prostate cancer. In my judgment, although the difference was statistically significant, its validity is suspect. This situation might be rectified by a longer median follow-up.

Guru Sonpavde, M.D.
Maricopa County Medical Center, Phoenix, AZ 85008

To the Editor:

The definition of death due to prostate cancer used by Holmberg and colleagues is problematic. Death was attributed to prostate cancer “if the autopsy determined that death was due to prostate cancer or there were distant metastases that had progressed or had not responded to treatment.” The use of this definition may have spuriously increased mortality due to prostate cancer, since men who died with advanced prostate cancer but not from it would have been categorized as having died from the disease. Since metastatic disease was more common in the watchful-waiting group, this problem would have been accentuated in this group and may account for the lower observed mortality attributable to causes other than prostate cancer in the watchful-waiting group than in the prostatectomy group. The authors noted six fewer deaths from other causes in the watchful-waiting group, but there is no a priori reason that mortality from other causes should have been lower in that group. If the number of deaths from prostate cancer were decreased by six in the watchful-waiting group, the mortality attributable to prostate cancer would be 7.2 percent in this group, making the difference between watchful waiting and radical prostatectomy no longer statistically significant (P=0.14).

Scott D. Stern, M.D.
University of Chicago, Chicago, IL 60637
sstern@medicine.bsd.uchicago.edu

Author/Editor Response

We agree with Blanker et al. that a per-protocol analysis might give additional insight into the effects of surgery, but we chose to follow our plan to put our hypothesis to a rigorous intention-to-treat test. The monitoring process did not show that the indications for the use of skeletal scintigraphy or prostate-specific–antigen testing were different in the groups.

Only about 5 percent of the men in our study had a Gleason score of more than 7, so our trial will never be powered to answer the question raised by Drs. Gambacorti-Passerini and Sala. Other studies are needed to determine whether the Gleason score modifies the effect of surgery.

Dr. Sonpavde is right that longer follow-up is necessary, and we plan to continue follow-up for a long time to come. Dr. Stern's question about bias due to our definition of causes of death will be clarified when the study has greater power for analyses of both mortality from prostate cancer and mortality from any cause. However, it is unlikely that such a major bias was introduced into our study, because it would require that men with a distant recurrence have only a small or moderate risk of dying from prostate cancer. That phenomenon is not usual in the natural history of prostate cancer. With one exception, all the men who died from prostate cancer in our study had clinically verified progressive, distant disease before death.

Anna Bill-Axelson, M.D.
Lars Holmberg, M.D., Ph.D.
University Hospital, SE-751 85 Uppsala, Sweden
anna.bill-axelson@urologi.uas.lul.se

Citing Articles (2)

Citing Articles

1. 1

   Sadao Kamidono, Shinichi Ohshima, Yoshihiko Hirao, Kazuhiro Suzuki, Yoichi Arai, Hiroyuki Fujimoto, Shin Egawa, Hideyuki Akaza, Isao Hara, Shiro Hinotsu, Yoshiyuki Kakehi, Tomonori Hasegawa. (2008) Evidence-based Clinical Practice Guidelines for Prostate Cancer (Summary - JUA 2006 Edition). International Journal of Urology 15:1, 1-18
   CrossRef

2. 2

   Nikki Peters, Katrina Armstrong. (2005) Racial Differences in Prostate Cancer Treatment Outcomes. Cancer Nursing 28:2, 108???118
   CrossRef