Measurable disease is defined as a baseline difference between involved (amyloidogenic) and uninvolved free light chains >50 mg/L. Forty subjects with measurable disease died before evaluation of response and are considered nonresponders in the present intent-to-treat analysis. Data on bortezomib dosage were available in 196 of 201 patients with measurable disease. Cardiac staging is based on NT-proBNP (cutoff, 332 ng/L) and cTnT (cutoff, 0.035 ng/mL), or cTnI (cutoff, 0.1 ng/mL), with stage I, II, and III patients having none, one, or two markers above the cutoff, respectively. Stage IIIa patients have NT-proBNP ≤8500 ng/L and stage IIIb patients have NT-proBNP >8500 ng/L. Bortezomib dosage: full dose, 1.3 mg/m² twice weekly or 1.6 mg/m² once weekly; intermediate dose, 1.0 mg/m² twice weekly or 1.3 mg/m² once weekly; low dose, <1.0 mg/m² twice weekly or 1.3 mg/m² once weekly. Dexamethasone dosage: full dose, at least 160 mg per cycle; intermediate dose, <160 mg and ≥80 mg per cycle; low dose, <80 mg per cycle.

PR, partial response.

P < .05 compared with stages I, II, and IIIa.

P < .05 compared with full dose. Among the 15 patients who received 1.0 mg/m² twice weekly, 2 achieved VGPR and 4 reached PR. Notably, in a multiple logistic regression analysis, only cardiac stage IIIb (P = .008) and not low doses of bortezomib (P = .191) and dexamethasone (P = .353), was an independent predictor of hematologic response, indicating that lower response rates in patients receiving lower doses of bortezomib or dexamethasone were likely explained by early deaths. Response rate was not affected by twice or once weekly administration (67% vs 62%; P = .549), or by IV or subcutaneous bortezomib administration (63% vs 61%; P = .862).