Reply to K. Lin et al

Abstract

We thank Lin et al for their interest and for their thoughtful comments on our article. They question the validity of our analysis, indicating a potential bias because of the exclusion of patients who discontinued hormonal therapy as a result of disease progression and death from causes other than prostate cancer. We agree with their point, and we did address the limitations of our hypothesis-generating analysis, including this issue, in our article. To overcome the specific limitation mentioned by Lin et al and to address the issue of bias, we performed a landmark analysis looking only at those patients who survived beyond 5 years in all three groups. The results of this analysis have been published in a previous correspondence to Journal of Clinical Oncology. Briefly, of the total of 189 patients who received less than 5 years of hormonal therapy in the original analysis, 145 (77%) were alive for at least 5 years. There were no significant changes in the distribution of pretreatment characteristics for the smaller, landmark subset of patients compared with the original 189-patient cohort. The 5-year landmark analysis continues to show a trend for an overall survival difference favoring more than 5 years hormonal duration (11-year survival, 64% v 50%), although this difference is no longer statistically significant (P .2), perhaps because of the smaller sample. There are, however, still statistically significant differences among the three hormonal therapy duration groups with respect to disease-free survival (P .0009), local failure (P .02), and distant metastases (P .002), favoring the longer hormonal therapy duration. The optimal duration of adjuvant hormonal therapy for patients with high-risk prostate cancer remains undetermined. As stated by Lin et al, a recent randomized trial shows a detrimental effect on survival for a shorter course of hormonal therapy. Whether treatment duration beyond 5 years will be associated with a significant improvement in survival remains to be established, and only a properly designed randomized trial will unequivocally answer this provocative question.