TAX-327: Docetaxel/prednisone versus mitoxantrone/prednisone

Study design

The patients enrolled in this trial had hormonerefractory metastatic prostate cancer and a testosterone level in the castrate range. Patients were allowed to have received only one prior chemotherapy treatment with estramustine and were withdrawn from antiandrogen therapy. The trial’s endpoint was survival. We wanted to detect a hazard ratio of 0.75 for survival in favor of docetaxel (Tannock 2004).

The three treatment arms included: (1) docetaxel 75 mg/m² every three weeks plus prednisone, (2) docetaxel 30 mg/m² weekly for five out of six weeks plus prednisone and (3) mitoxantrone 12 mg/m² every three weeks plus prednisone (4.1). We enrolled 1,006 patients over two years, and the analysis occurred about three and a half years after the first patient was enrolled. Each treatment arm had more than 300 patients.

Efficacy

With a median follow-up of about 20.7 months, the median survival for patients treated with every three-week docetaxel and prednisone was 18.9 months, compared to a median survival of 16.5 months for those treated with mitoxantrone and prednisone. Forty-five percent and 48 percent of patients treated with every three-week and weekly docetaxel had a 50 percent decline in their PSA that lasted for at least three weeks, and 32 percent of the patients treated with mitoxantrone and prednisone had a 50 percent decline in their PSA, which was significantly different (p < 0.001) (Tannock 2004; [1.3]).

About 30 percent of the patients in the docetaxel arms had a reduction in pain, compared to about 20 percent of the patients treated with mitoxantrone and prednisone. The difference in the reduction in pain between mitoxantrone plus prednisone and every three-week docetaxel plus prednisone was also significant (p = 0.01). Very few objective responses in soft tissue metastases were reported in all three arms (Tannock 2004).

Adverse events

The toxicity was as predicted with these compounds, mostly myelosuppression. Thirty-two percent of the patients treated with every three-week docetaxel and prednisone had myelosuppression (Grade III/IV neutropenia), but less than three percent had neutropenic fever, documented sepsis or death. Only 1.5 percent of the patients receiving weekly docetaxel and prednisone had myelosuppression (Grade III/IV neutropenia), compared to about 20 percent of the patients on mitoxantrone and prednisone. The incidence of febrile complications was very low and similar in all three treatment arms (Eisenberger 2004; [4.2]).

The other toxicities were minor (≤Grade II) and not dose limiting. There was some neuropathy, fatigue and edema in the patients treated with docetaxel, which is more toxic than mitoxantrone and prednisone, but the toxicities were quite reasonable. We had very few episodes of significant nausea and vomiting and some changes in liver function tests. Alopecia was reported more frequently with every three-week docetaxel, and changes in the nails and eyes were reported more with weekly docetaxel (Eisenberger 2004; [4.2]). About 16 percent of the patients on weekly docetaxel discontinued treatment because of an adverse drug reaction (Tannock 2004), compared to only 11 percent on every three-week docetaxel.

Comparing the results from SWOG-S9916 to those from TAX-327

The results from those two trials are very similar. In SWOG-S9916, survival for the patients treated with docetaxel plus estramustine was 17.5 months compared to 15.6 months for those on mitoxantrone plus prednisone. A PSA response occurred in 50 percent of the patients on docetaxel plus estramustine, compared to 27 percent of those on mitoxantrone plus prednisone (Petrylak 2004; [1.3]).

The difference between the two trials was in the toxicities. Although no head-to- head comparison was conducted, estramustine plus docetaxel was more toxic than docetaxel plus prednisone. The most significant toxicities were cardiovascular or thrombotic. Halfway through SWOG-S9916, the protocol was amended to include prophylactic anticoagulation (ie, warfarin plus aspirin) in the patients treated with estramustine (Petrylak 2004a).

Select publications

Dr Eisenberger is the R Dale Hughes Professor of Oncology and Urology at The Johns Hopkins University in Baltimore, Maryland.