Prognostic Value Of sFLC Ratio At Baseline On Response After Induction Therapy In Patients With Multiple Myeloma In The GMMG MM5 Trial

Studies in newly diagnosed multiple myeloma (MM) described a prognostic value of the serum free light chain ratio (sFLCr) on progression free survival (PFS) and overall survival (OS). The GMMG MM5 phase III trial compares VCD (bortezomib, cyclophosphamide, dexamethasone) and PAd (bortezomib, adriamycin, dose-reduced dexamethasone) for induction therapy followed by stem cell mobilization and harvest, high-dose therapy and a lenalidomide-based consolidation/maintenance therapy for 2 years vs. lenalidomide until complete response (CR). First primary end point of the ongoing study was response after induction, second primary end point was progression-free survival. To analyse if the sFLCr has a prognostic value on the response after induction therapy with respect to achieve a very good partial response (VGPR) or better, we investigated patients with newly diagnosed MM which were included in the MM5 trial.

504 patients between 18-70 years with newly diagnosed MM were included in the MM5 trial between July 2010 and October 2012. Patient serum samples were collected and sent to a central laboratory for analysis of the free light chains (sFLC) in serum prior to treatment and after induction therapy. For the sFLC measurement the Freelite^® assay (The Binding Site) was used.

The analysis was based on the intention-to-treat (ITT) population (502 patients evaluable). The sFLC ratio (sFLCr) was available for n=498 patients at baseline. After induction treatment, there were 179 patients achieving CR and VGPR (VGPR+ group) and 318 patients not achieving at least VGPR (PR- group). There was no significant difference in the response between the PAd and the VCD (reported separately), so we analysed the data irrespective of the treatment arm.

We examined in univariate and multivariate analyses the association of response VGPR+ after induction treatment with clinical variables (age, gender, ISS, CRAB criteria, IgA vs non-IgA type MM, WHO status), LDH measurement, cytogenetic factors (deletion 17p13, gain 1q21, translocation (4;14) ) and sFLCr at baseline. sFLCr was analysed either as categorized factor for values within the range [1/32, 32] or outside this range, as proposed by Snozek et al. (2008), or as continuous factor, where the absolute value of log2 transformed sFLCr is used for analyses. Further, LDH values were log2 transformed for analyses. In univariate analysis, the groups of VGPR+ and PR- showed no significant difference in age, gender, ISS, presence of anemia or bone disease, WHO state or the cytogenetic abnormalities del17p13, gain 1q21 or translocation t(4;14), but were significantly associated with elevated Calcium (p=0.07), renal insufficiency (p=0.004), presence of IgA type MM (p<0.001) and LDH measurement (p<0.001). There was no significant difference between patients with sFLCr at baseline within or outside the range set at 1/32 – 32 in achieving VGPR+ (p=0.23) whereas the absolute value of log2 transformed values was significantly associated with response (p=0.02).

VGPR+ was further analysed by multivariate logistic regression models which were adjusted for the clinical and cytogenetic variables, for LDH and for sFLCr (either as categorical or continuous factor). sFLCr revealed as significant prognostic factor when considered as continuous variable (p = 0.03) but failed significance as categorized factor. Patients with extreme sFLCr values showed a better chance of VGPR+, as did IgA myeloma patients (p=0.03) and patients with renal insufficiency (p=0.04).

Cut point analyses showed that in our trial patients with a baseline sFLCr < 1/1097 and >1097 had a significantly better chance to achieve VGPR+ after induction compared to patients with a sFLCr within this range. There were 44 patients below the range (IgA 22.7%, IgG 34.1%, IgD 6.8%, light chain MM (LCMM) 36.4%), 412 patients in the range (IgA 20.9%,IgG 63.8%, IgD 1%, IgM 0.2%, LCMM 14.1%) and 42 patients above the range (IgA 19.1%, IgG 42.9%, LCMM 38.1%).

In conclusion, the cut point for baseline sFLCr published by Snozek et al. did not define patient groups with significantly different response in the MM5 trial. This cut point was set for the outcome OS and PFS and could not be replicated for response after induction treatment in our trial. The cut point which showed a significant difference in achieving VGPR+ was much higher in our population which might be explained by the higher proportion of light chain MM below and above the cut point range.