As of August 04, 104 of 668 patients enrolled in TT 2 have relapsed/progressed, 61 of whom had been randomized to no THAL and 43 to THAL. The 2 groups were comparable in 14 pre-treatment features examined including the presence of cytogenetic abnormalities (CA), LDH, B2M and albumin. At the time of treatment failure, which occurred at a similar median time from study enrollment of 23 months, higher proportions of patients on THAL had CA (63% vs 37%, p=.03) and higher levels of bone marrow plasmacytosis >33% (46% vs 12%, p=.001). Their subsequent management included THAL + DEX in 25%; Velcade + THAL in 16% and DT-PACE in 10% of those failing on the no THAL arm; compared to THAL + DEX in 12%; Velcade + THAL in 14% and DT-PACE in 21% of those relapsing on the THAL arm. Subsequent salvage response was 51% for those relapsing on no THAL and 26% of those on THAL (relapsing on THAL) (p=.028). Importantly, survival from the time of TT 2 failure was significantly better when patients had not received prior THAL than when THAL was part of the initial management (median of 29 vs 8 mos, p=.0001).

Univarately adverse features for post-failure survival included the presence of CA at the time of failure, IgA isotype, LDH at baseline prior to TT 2 and initial randomization to THAL. On multivariate analysis, baseline LDH (HR 2.6, p=.0006) and THAL randomization (HR 2.6, p=.0006) were highly associated with inferior post-TT 2 failure survival (R2 =38%). Indeed, those on the THAL arm had inferior survival regardless of CA, present either at baseline or at relapse (Figure 1). Our preliminary conclusion is that initial use of THAL for the primary management of symptomatic MM reduces the subsequent salvage potential, due to the emergence of a more aggressive disease exhibiting CA at relapse and greater tumor burden (high bone marrow plasmacytosis). Thus, randomized trials are needed to determine whether thalidomide is more appropriately used up front or for the management of relapse in order to achieve the longest overall survival.

Author notes

Corresponding author

Sign in via your Institution