Correspondence

Radical Prostatectomy versus Watchful Waiting

N Engl J Med 2005; 353:1298-1300September 22, 2005

Article

To the Editor:

Bill-Axelson et al. (May 12 issue)1 state that disease-specific mortality after 10 years of follow-up was reduced by 5.3 percentage points among men assigned to radical prostatectomy, favoring radical prostatectomy over watchful waiting. However, from the data in Table 3 of their article, it appears that 35 subjects were lost from the prostatectomy group and 12 subjects were lost from the watchful-waiting group. If the numbers for disease-specific mortality (30 deaths in the radical-prostatectomy group and 50 deaths in the watchful-waiting group) are used to recalculate the disease-specific mortality, and if it is assumed that all missing subjects in the prostatectomy group died and even that all missing subjects in the watchful-waiting group died (the worse-case scenario for both groups), the results do not favor prostatectomy over watchful waiting: in the prostatectomy group, disease-specific mortality would be reduced by 0.7 percentage point (P=0.8).

Michael E. Stuart, M.D.
Sheri A. Strite, B.A.
Delfini Group, Seattle, WA 98115
mstuart@delfini.org

1 References

1. 1

   Bill-Axelson A, Holmberg L, Ruutu M, et al. Radical prostatectomy versus watchful waiting in early prostate cancer. N Engl J Med 2005;352:1977-1984
   Full Text | Web of Science | Medline

To the Editor:

In the study by Bill-Axelson et al., there was an inconsistency in the statistical results for absolute risk reduction versus relative risk. For 10-year disease-specific mortality (Table 3 of the article), the relative risk with radical prostatectomy is 0.56 (95 percent confidence interval, 0.36 to 0.88), whereas the absolute risk reduction is 5.3 percent (95 percent confidence interval, –0.3 percent to 11.0 percent). The relative risk is statistically significant and the absolute risk reduction is not. Since the upper bound of the confidence interval for the relative risk is quite far from the null value of 1, it is hard to understand how the absolute risk reduction could not be statistically significant.

Paul R. Marantz, M.D., M.P.H.
Charles B. Hall, Ph.D.
Carol A. Derby, Ph.D.
Albert Einstein College of Medicine, Bronx, NY 10461

To the Editor:

Decisions about undergoing radical prostatectomy can be influenced by men's fears of sexual dysfunction or urinary incontinence, which are common complications of this surgery. Thus, the lack of any mention of these outcomes in the study by Bill-Axelson et al. is surprising. The same men who underwent surgical intervention in this study had significantly more erectile dysfunction (80 percent vs. 45 percent) and urinary leakage (49 percent vs. 21 percent) than the men in the watchful-waiting group, as described in a previous publication.1 Follow-up of these important outcomes is notably absent, as is any reference to impotence or incontinence. Instead, the authors speculate that the men in the watchful-waiting group might actually have decreased well-being and quality of life.

Hillary K. Liss, M.D.
Joann G. Elmore, M.D., M.P.H.
University of Washington School of Medicine, Seattle, WA 98104

1 References

1. 1

   Steineck G, Helgesen F, Adolfsson J, et al. Quality of life after radical prostatectomy or watchful waiting. N Engl J Med 2002;347:790-796
   Full Text | Web of Science | Medline

To the Editor:

Bill-Axelson et al. claim to include patients with early prostate cancer, but most of the patients appear to have had locally advanced disease. Approximately 75 percent of patients had T2 disease, and almost 50 percent had a prostate-specific antigen level of greater than 10 ng per milliliter. Modern nomograms indicate that a large proportion of these patients will have extracapsular invasion.1

In addition, approximately 5 percent of patients had a Gleason score of between 8 and 10, since the eligibility criteria included only those patients with well-differentiated or moderately well-differentiated prostate cancer. Therefore, the tumors may be characterized as of relatively low grade but bulky.

Guru Sonpavde, M.D.
Texas Oncology, Houston, TX 77598
guru.sonpavde@usoncology.com

1 References

1. 1

   Partin AW, Mangold LA, Lamm DM, Walsh PC, Epstein JI, Pearson JD. Contemporary update of prostate cancer staging nomograms (Partin Tables) for the new millennium. Urology 2001;58:843-848
   CrossRef | Web of Science | Medline

Author/Editor Response

Stuart and Strite have misinterpreted our data in Table 3 and the principles of censoring in statistical analysis. No subjects were lost to follow-up in our study, and all calculations were done according to the intention-to-treat principle.

The statistical results for absolute risk reduction and relative risk are not inconsistent, nor are they unusual, as argued by Marantz et al. Statistical significance is test-dependent — a fact that underscores the problems of relying heavily on statistical significance rather than on quantitative estimation and biologic reasoning in interpreting the outcomes of clinical studies. For this study, we applied the convention that is most often used to test the relative risk reduction. Furthermore, the lower band of the absolute risk reduction is very close to zero, and a log-rank test yields a P value that is very similar to that of Gray's test.

Liss and Elmore are correct in that follow-up of symptoms and quality of life is very important. Our speculation about well-being and quality of life in the watchful-waiting group did not come out of the blue, since the symptom burden is correlated to well-being and quality of life. In our study, the symptom burden increases in the watchful-waiting group, especially among the younger men.

We agree with Sonpavde that because we do not know exactly how our results translate to the early cancers detected today, our results present a difficulty in clinical decision making, as we pointed out in our paper. However, it is certain that even if a better relative risk reduction than reported in our study can be achieved, the resulting absolute reduction in a group that enjoys a very good prognosis at the outset will be moderate or small.

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

Jan-Erik Johansson, M.D., Ph.D.
Örebro University Hospital, SE-701 55 Örebro, Sweden

Citing Articles (1)

Citing Articles

1. 1

   Andrea Gallina, Nazareno Suardi, Francesco Montorsi, Umberto Capitanio, Claudio Jeldres, Fred Saad, Markus Graefen, Shahrokh F. Shariat, Hugues Widmer, Philippe Arjane, François Péloquin, Paul Perrotte, Pierre I. Karakiewicz. (2008) Mortality at 120 days after prostatic biopsy: A population-based study of 22,175 men. International Journal of Cancer 123:3, 647-652
   CrossRef