Skip to main content
. 2012 Jun 6;120(8):1597–1600. doi: 10.1182/blood-2012-04-421883

Table 4.

Univariate Cox proportional hazards regression modeling time from enrollment to SPMs

Variable n/N (%) Time to SPMs from enrollment
HR (95% CI) P*
Female sex 451/1148 (39%) 0.77 (0.47-1.25) .288
White race 1008/1148 (88%) 2.65 (0.97-7.25) .058
Age ≥ 65 y 266/1148 (23%) 2.06 (1.25-3.39) .004
IgA isotype 270/1143 (24%) 0.83 (0.46-1.48) .524
IgG isotype 624/1143 (55%) 0.85 (0.54-1.34) .487
Albumin < 3.5 g/dL 275/1143 (24%) 0.66 (0.33-1.33) .249
B2M ≥ 3.5 mg/L 478/1145 (42%) 1.03 (0.64-1.68) .895
B2M > 5.5 mg/L 238/1145 (21%) 1.43 (0.80-2.56) .233
IPSS Stage I 565/1141 (50%) 1.24 (0.76-2.00) .392
IPSS Stage II 339/1141 (30%) 0.60 (0.33-1.10) .098
IPSS Stage III 238/1145 (21%) 1.43 (0.80-2.56) .233
CRP ≥ 8 mg/L 423/1136 (37%) 0.85 (0.51-1.39) .512
Creatinine ≥ 2.0 mg/dL 96/1133 (8%) 1.81 (0.87-3.79) .113
Hb < 10 g/dL 312/1146 (27%) 0.90 (0.52-1.57) .716
LDH ≥ 190 U/L 328/1145 (29%) 1.30 (0.78-2.16) .309
Platelet count < 150 × 109/L 171/1146 (15%) 0.53 (0.22-1.33) .177
Cytogenetic abnormalities 367/1136 (32%) 1.64 (1.01-2.66) .045
GEP70 high-risk 123/792 (16%) 0.71 (0.22-2.30) .571
GEP80 high-risk 56/792 (7%) 0.68 (0.09-4.92) .701
K/L FLC ratio N = 478 1.00 (1.00-1.00) .515
Kappa FLC N = 614 1.00 (1.00-1.00) .464
Kappa FLC > 168 70/614 (11%) 1.23 (0.43-3.51) .697
Lambda FLC N = 615 1.00 (1.00-1.00) .456
Thalidomide 803/1148 (70%) 1.16 (0.71-1.89) .561

HR indicates hazard ratio; 95% CI, 95% confidence interval; B2M, β-2-microglobulin; IPSS, International Prognostic Scoring System; CRP, C-reactive protein; Hb, hemoglobin; LDH, lactate dehydrogenase; GEP, gene-expression profiling; and FLC, free light chain.

*

P value from Wald χ2 test in cox regression. All univariate P values are reported regardless of significance. The multivariate model uses stepwise selection with an entry level of P = .1 and the variable remains if it meets the P = .05 level. A multivariate P value > .05 indicates a variable forced into the model with significant variables chosen using stepwise selection.