A randomized trial compared prednisone plus IV melphalan to prednisone plus oral melphalan in the treatment of myeloma. As discussed below, overall response rates at week 22 were comparable; however, because of changes in trial design, conclusions as to the relative activity of the 2 formulations after week 22 are impossible to make.
Both arms received oral prednisone starting at 0.8 mg/kg/day with doses tapered over 6 weeks. Melphalan doses in each arm were: Arm 1: Oral melphalan 0.15 mg/kg/day x 7 followed by 0.05 mg/kg/day when WBC began to rise. Arm 2: IV melphalan 16 mg/m2 q 2 weeks x 4 (over 6 weeks) followed by the same dose every 4 weeks. Doses of melphalan were adjusted according to the following criteria:
Table 1. Criteria for Dosage Adjustment in a Randomized Clinical Trial
| WBC/mm3 | Platelets | Percent of Full Dose |
| ≥4,000 | ≥100,000 | 100 |
| ≥3,000 | ≥75,000 | 75 |
| ≥2,000 | ≥50,000 | 50 |
| <2,000 | <50,000 | 0 |
One hundred seven patients were randomized to the oral melphalan arm and 203 patients to the IV melphalan arm. More patients had a poor-risk classification (58% versus 44%) and high tumor load (51% versus 34%) on the oral compared to the IV arm (P<0.04). Response rates at week 22 are shown in the following table:
Table 2. Response Rates at Week 22
| Initial Arm | Evaluable Patients | Responders n (%) | P |
| Oral melphalan | 100 | 44 (44%) | P>0.2 |
| IV melphalan | 195 | 74 (38%) |
Because of changes in protocol design after week 22, other efficacy parameters such as response duration and survival cannot be compared.
Severe myelotoxicity (WBC ≤1,000 and/or platelets ≤25,000) was more common in the IV melphalan arm (28%) than in the oral melphalan arm (11%).
An association was noted between poor renal function and myelosuppression; consequently, an amendment to the protocol required a 50% reduction in IV melphalan dose if the BUN was ≥30 mg/dL. The rate of severe leukopenia in the IV arm in the patients with BUN over 30 mg/dL decreased from 50% (8/16) before protocol amendment to 11% (3/28) (P = 0.01) after the amendment.
Before the dosing amendment, there was a 10% (8/77) incidence of drug-related death in the IV arm. After the dosing amendment, this incidence was 3% (3/108). This compares to an overall 1% (1/100) incidence of drug-related death in the oral arm.