Work

Heavy/light chain ratio normalization prior to transplant is of independent prognostic significance in multiple myeloma: a BMT CTN 0102 correlative study

Year: 2016

Type: article

Abstract: The heavy light chain (HLC) immunoassay quantifies light chain types of each immunoglobulin (Ig), enabling the calculation of ratios of HLC pairs (Katzmann & Rajkumar, 2013). This assay measures the light chain types of each Ig (G/A/K), generating ratios of monoclonal Ig/background polyclonal Ig concentrations as paired ratios. It has shown clinical utility in multiple myeloma (MM), particularly for recognition of small M-proteins that migrate in the beta or gamma fraction among polyclonal immunoglobulins, for monitoring of minimal residual disease and as a predictor of progression of monoclonal gammopathy of undetermined significance to MM (Bradwell et al, 2013; Katzmann et al, 2013; Ludwig et al, 2013; Boyle et al, 2014). Additionally, the HLC assay may be a surrogate marker of normal plasma cell recovery after transplant (Tovar et al, 2012). We report the results of a correlative study testing the additional prognostic role of a normalized HLC and free light chain (FLC) tests over conventional electrophoretic response in predicting outcomes done on archived samples of patients enrolled in Blood Marrow Transplant Clinical Trials Network (BMTCTN) 0102, a randomized phase 3 multicentre trial aimed at determining the role of tandem autologous or allogeneic second transplant after upfront autologous haematopoietic cell transplant (HCT) in MM (Krishnan et al, 2011). Eligible patients were biologically assigned to receive a tandem allogeneic (alloHCT) or a second autologous (autoHCT) transplant, based on availability of a human leucocyte antigen (HLA)-matched sibling donor. The tandem autoHCT group was re-randomized to receive 1 year of maintenance therapy with thalidomide plus dexamethasone versus observation. De-identified, archived cryopreserved serum samples collected prior to first transplant, day 56 and 1 year post-second transplant banked at the National Heart Lung and Blood Institute (NHLBI) sample biorepository were used in this study using an Institutional Review Board-approved protocol. Samples were processed at The Binding Site, Birmingham, UK. Testing was performed using a Siemens BNII nephelometer for free kappa and free lambda (FLC Freelite® kits), IgGκ and IgGλ, IgAκ and IgAλ, IgMκ and IgMλ (HLC IgG, IgA and IgM Hevylite® kits). The normal ranges for ratios of the FLC and HLC assays were determined by analysis of serum samples from 146 normal controls and were: FLC κ/λ 0·26–1·65; HLC IgG κ/λ 0·98–2·75; HLC IgA κ/λ 0·8–2·04 and IgM κ/λ 0·96–2·3. Twenty-eight samples could not be analysed due to inadequate sample collection; 1 patient with disease progression at study entry (auto-auto arm) was excluded; 490 samples were analysed. The International Uniform Response Criteria (Durie et al, 2006) was used to define disease response with the addition of near complete response (CR), which was defined as evidence of disease by immunofixation electrophoresis without morphological evidence of MM in bone marrow.(Krishnan et al, 2011) Normalization of HLC ratios across all 3 measured heavy/light chain pairs or normal clonal isotype with normal ratios of uninvolved pairs was considered HLC-R while normalization of the FLC ratio was called FLC remission (FLC-R). The HLC result of any HLC pair missing for a given time was disregarded, thus decreasing the number of interpretable HLC compared to FLC ratios (Table 1). Sensitivity and specificity of FLC-R and HLC-R were calculated compared to electrophoretic response. Progression-free survival (PFS) was defined from first registration to the first observation of progression. Multivariate models were used to identify baseline variables impacting PFS and overall survival (OS) after transplant and adjusted models were generated to identify the prognostic value of pre-transplant HLC-R and FLC-R for those in electrophoretic CR or very good partial response (VGPR). Model building was performed by Cox regression and survival was estimated by the Kaplan–Meier method. Statistical analyses were performed using SAS version 9 (SAS Institute, Cary, NC, USA). Table 1 shows the baseline characteristics and test results. The alloHCT group was younger (median age at diagnosis 51 years vs. 56 years, P < 0·001), more frequently white (82% vs. 74%, P = 0·03) and with high-risk disease (18% vs. 10%, P = 0·01) compared to the autoHCT group, similar to the demographics reported in the parent study (Krishnan et al, 2011). The median follow-up was similar in both groups. Heavy light chain remission had 100% sensitivity and 100% negative predictive value for identifying electrophoretic CR and non-CR respectively, i.e. a normal HLC result was always concordant with electrophoretic CR while an abnormal HLC result was always concordant with electrophoretic non-CR. Further, HLC-R had high sensitivity (89%) and low specificity (52%) to distinguish ≥VGPR from <VGPR, whereas FLC-R showed high specificity (81%) with low sensitivity (49%) between these groups. Only 23 patients achieving ≥VGPR had HLC-NR, thus restricting our ability to identify differences in outcomes. While HLC response did not help separate patients with a VGPR or better (Fig 1A), those patients with pre-transplant <VGPR [partial response (PR) or stable disease (SD)] status and HLC-R had better outcomes than those without HLC-R (P = 0·03; Fig 1B). It is known that patients with pre-transplant electrophoretic ≥VGPR have good post-transplant outcomes overall (Attal et al, 1996), thus it is not surprising that the HLC had no additional benefit among these patients. Ludwig et al (2013) showed similar results, with improvements in survival among a small group of patients stratified by HLC ratio, with benefit seen particularly in the group with PR rather than ≥VGPR. Multivariate analysis associated patients with ≥VGPR response prior to first transplant and Durie-Salmon stage (DSS) I/II versus III with superior PFS and OS. After adjusting for electrophoretic response, DSS and study arm (autoHCT versus alloHCT), HLC-R was an independent predictor of superior PFS [hazard ratio (HR) 0·75, 95% confidence interval (CI) 0·59–0·95, P = 0·016;Fig 1C], freedom from relapse (HR 0·77, 95% CI 0·6–0·98, P = 0·035) and OS (HR 0·69, 95% CI 0·51–0·92, P = 0·012; Fig 1D]. No similar prognostic value was seen by incorporating FLC-R into the model, as also recently shown by Martinez-Lopez et al (2015). In conclusion, we show that HLC-R is 100% concordant to an electrophoretic CR, but an HLC-R can be achieved without establishment of an electrophoretic CR. The HLC-R status identifies a good prognostic risk group among patients who have not achieved a VGPR or better prior to transplant. For those in an electrophoretic CR, an FLC-R was not additionally prognostic. Our results indicate that HLC-R could be integrated into other response levels to further stratify patients not achieving CR/VGPR. Funding for this study was provided by The Binding Site, in support of BMT CTN Protocol 0201 which was funded by grant #U10HL069294 from the National Heart, Lung, and Blood Institute and the National Cancer Institute, the Department of the Navy, Office of Naval Research and the National Marrow Donor Program. Any views, opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views or the official policy or position of the above mentioned parties. All authors designed the study, collected and analysed the data, and wrote the manuscript; BL performed the statistical analysis and edited the article. All authors critically revised the manuscript for important intellectual content and approved the final draft. The authors have no conflicts of interest to report.

Source: British Journal of Haematology

DOI: 10.1111/bjh.14170