Androgen deprivation therapy (ADT) use in Medicare beneficiaries with nonmetastatic (M0) prostate cancer (PC) in the United States.

Abstract

e15169 Background: ADT is well-established for metastatic PC but is also used in men with less advanced PC. Patterns of ADT use in the M0 PC setting have not been well-described. Methods: Medicare claims in 2005 for men aged ≥66 years (yrs) were assessed for use of gonadotropin-releasing hormone agonists or bilateral orchiectomy. The cohort was limited to men who, during the 15 months (mos) prior to their first 2005 ADT claim, had continuous Parts A + B coverage, a diagnosis code for PC, no claims for metastases (except lymph nodes), and no claims for ADT. Follow-up (f/u) was from ADT initiation to 3 yrs, death, or change in Medicare coverage. Regimen duration was defined by combining claim service count and dosage with FDA label dosing. Because ADT has biologic activity beyond the recommended regimen duration, active dose time was calculated by adding 3 and 6 mos to claims for regimens <6 mos and ≥6 mos, respectively. Interruption was defined as ≥180 days from end of active dose time to next ADT claim. Results: We identified 3246 M0 PC patients aged ≥66 yrs who initiated ADT in 2005 but describe results in the 2161 who had 3 full yrs of f/u (data for the entire cohort were similar). Nearly 70% received only 1 type of ADT agent during f/u (25% had 2 types, 5% had 3 or 4 types), and 73% started on a 3-or 4-month regimen. The distribution of patients by first ADT type (and by type received at any point) was: leuprolide injection: 54% (67%); goserelin implant: 36% (43%); triptorelin injection: 6% (19%); leuprolide implant: 2% (4%); orchiectomy: <1% (1%); and histrelin implant: 0% (3%). Use of >1 type of ADT agent was more common in men who began with goserelin (45%) or triptorelin (53%) versus leuprolide (≤24%). At three years after ADT initiation, 36% received ADT for the entire 36-month period, 15% for 24-35 mos, 17% for 12-23 mos, and 32% for <12 mos. Interruption in active therapy occurred in only 13%. Conclusions: Most elderly men with M0 PC initiating ADT in 2005 started on a 3- or 4-month leuprolide or goserelin regimen. Though these agents remained the most commonly used throughout f/u, switching was common (30% used >1 type during f/u). ADT was continued for at least 2 yrs in nearly half and extended to 3 yrs in over one-third of the cohort.