The presence of high levels of XBP1s is associated with a poor OS. (A) Patients with XBP1s/u less than 1.33 (H indicates high XBP1s/u; L, ow XBP1s/u) have a longer OS (P = .03). (B) The presence of high levels of XBP1s has a significant effect on OS in patients with high β2M (P < .05; group 1: XBP1s/u > 1.33, β2M > 4, n = 75; group 2, XBP1s/u > 1.33, β2M < 4, n = 62; group 3: XBP1s/u < 1.33, β2M > 4, n = 34; group 4, XBP1s/u < 1.33, β2M < 4, n = 16). (C) The combination of the 3 independent risk factors XBP1s/u, β2M, and t(4;14) splits patients into 4 groups with significant differences on OS [P < .001; group 0 = no factor (n = 11); group 1 = 1 factor (n = 76); group 2 = 2 factors (n = 79); group 3 = with all 3 factors (n = 9)]. (D) Patients with a high XBP1s/u (n = 107) treated with thalidomide have a significantly shorter OS compared with patients with a low XBP1s/u (n = 29) treated in a similar way (P = .001). In addition, a significant effect in OS is observed comparing patients with low XBP1s/u who received thalidomide (n = 29) to those with no thalidomide treatment (n = 33; P = .001). (E) In 218 patients who have relapsed, those patients with a high XBP1s/u (n = 94) treated with thalidomide have a significantly shorter survival after relapse than patients with a low XBP1s/u (n = 24) treated in a similar way (P = .002). In addition, a significant effect in survival after relapse is observed when patients with low XBP1s/u who received thalidomide (n = 24) were compared with patients with no thalidomide treatment (n = 30; P = .01).