mg/m2 by short IV infusion every 3 weeks. Prednisone was administered orally at a dose of 5 mg twice a day. Patients randomized to the prednisone arm were crossed over to the N + P arm if they progressed or if they were not improved after a minimum of 6 weeks of therapy with prednisone alone.
A total of 161 patients were randomized, 80 to the N + P arm and 81 to the P arm. The median Novantrone dose administered was 12 mg/m2 per cycle. The median cumulative Novantrone dose administered was 73 mg/m2 (range of 12 to 212 mg/m2).
A primary palliative response (defined as a 2-point decrease in pain intensity in a 6-point pain scale, associated with stable analgesic use, and lasting a minimum of 6 weeks) was achieved in 29% of patients randomized to N + P compared to 12% of patients randomized to P alone (p = 0.011). Two responders left the study after meeting primary response criterion for two consecutive cycles. For the purposes of this analysis, these two patients were assigned a response duration of zero days. A secondary palliative response was defined as a 50% or greater decrease in analgesic use, associated with stable pain intensity, and lasting a minimum of 6 weeks. An overall palliative response (defined as primary plus secondary responses) was achieved in 38% of patients randomized to N + P compared to 21% of patients randomized to P (p = 0.025).
The median duration of primary palliative response for patients randomized to N + P was 7.6 months compared to 2.1 months for patients randomized to P alone (p = 0.0009). The median duration of overall palliative response for patients randomized to N + P was 5.6 months compared to 1.9 months for patients randomized to P alone (p = 0.0004).
Time to progression was defined as a 1-point increase in pain intensity, or a > 25% increase in analgesic use, or evidence of disease progression on radiographic studies, or requirement for radiotherapy. The median time to progression for all patients randomized to N + P was 4.4 months compared to 2.3 months for all patients randomized to P alone (p = 0.0001). Median time to death was 11.3 months for all patients on the N + P arm compared to 10.8 months for all patients on P alone (p = 0.2324).
Forty-eight patients on the P arm crossed over to receive N + P. Of these, thirty patients had progressed on P, while 18 had stable disease on P. The median cycle of crossover was 5 cycles (range of 2 to 16 cycles). Time trends for pain intensity prior to crossover were significantly worse for patients who crossed over than for those who remained on P alone (p = 0.012). Nine patients (19%) demonstrated a palliative response on N + P after crossover. The median time to death for patients who crossed over to N + P was 12.7 months.
The clinical significance of a fall in prostate-specific antigen (PSA) concentrations after chemotherapy is unclear. On the CCI-NOV22 trial, a PSA fall of 50% or greater for two consecutive follow-up assessments after baseline was reported in 33% of all patients randomized to the N + P arm and 9% of all patients randomized to the P arm. These findings should be interpreted with caution since PSA responses were not defined prospectively. A number of patients were ineva luable for response, and there was an imbalance between treatment arms in the numbers of eva luable patients. In addition, PSA reduction did not correlate precisely with palliative response, the primary efficacy endpoint of thi
