Factors Influencing Outcomes of Relapse from Front-Line Autologous Stem Cell Transplant (ASCT) over a 14 Year Period: A Single Centre Experience

Background. Chemotherapy followed by ASCT is standard of care in fit transplant-eligible newly diagnosed patients with Multiple Myeloma (MM). Most patients, however, will relapse and require further treatment. With increasing choice of therapies at relapse, we need to understand the determinants of treatment outcomes for these patients, to aid management decisions and patient counselling.

Methods. This was a retrospective analysis of sequential patients who relapsed from upfront ASCT carried out between 2000-2014. PFS1 was defined as time from ASCT to 1st progression or death; PFS2, time from ASCT to 2nd progression or death and 2nd PFS, time from start of salvage regimen to 2nd progression or death. Post relapse survival (PRS) was measured from progression date and overall survival (OS) from ASCT date. Disease response, and adverse FISH were assigned according to IMWG. Survival was estimated using Kaplan- Meier curves and correlative analysis by Cox regression models.

Results. Of 474 patients undergoing ASCT (2000-2014), 277 relapsed with a median age of 58 years (range 29-70).The M:F ratio was 1.85:1. Isotypes identified were: IgG 57.0%, IgA 22.7%, light chain 15.9%. Median PFS1 from ASCT was 20 months (0.47-128.3). ORR pre-ASCT was 91.3% (sCR/CR 4.4%, VGPR 25.4%) and post ASCT was 95.3% (sCR/CR 14.5%, VGPR 52.2%). At relapse, median age was 60 years (30-73) and 36.5% of patients had ISS stage2-3 disease. For the 277 patients who relapsed, median OS from ASCT was 67 months (CI 57-73) and median PRS was 40 months (35-44). PFS1 had a strong influence on OS, HR 0.97 (95%CI: 0.96-0.97, p<0.001), and PRS HR 0.99 (0.98-0.996, p=0.004).

Patients relapsing before 2008 (prior to bortezomib funding) had shorter OS compared to those relapsing thereafter: median 61 vs 69 months, HR 1.38 (1.01-1.87, p=0.04). Depth of response (CR/VGPR) pre and post-ASCT were associated with longer PFS1- HR 0.70 (0.54-0.92, p=0.01), and HR 0.68(0.51-0.89, p=0.005) respectively, but not OS or PRS.

248 patients (89.5%) received systemic salvage therapy at relapse;106 patients (38.3%) experienced biochemical IMWG progression rather than clinical relapse, with median time to treatment of 5 months (2-64). Salvage regimens included: proteasome inhibitors(PIs) (64.5%), immunomodulatory agents (IMiDs, 29.8%, thalidomide-63.5%, lenalidomide-36.5%) and chemotherapy 5.2%; 26.6% of patients entered clinical trials and 13.7% underwent salvage ASCT. ORR was 70.4% (sCR/CR 10.9%, VGPR 31.6%). Median 2nd PFS was 17 months (16-20) and median PFS2 overall was 39 months (35-41). Achieving a deeper response to salvage treatment (CR/VGPR) was associated with a longer 2nd PFS (21 vs 17 months for PR, HR 0.65, 0.46-0.91, p=0.01), with a trend for PRS, HR 0.89(0.59-1.34, p=0.58) and OS, HR 0.77(0.51-1.17, p=0.22).

Novel agents induced deeper responses, CR/VGPR with PI regimens 51.3%, with IMiDs 30%, and with chemotherapy 8%. Patients treated with PIs compared to all other systemic treatment regimens had a significantly longer OS of 80 vs 48 months, HR 0.60(0.43-0.85, p= 0.004). Patients entered into clinical trials (66) had deeper responses, ≥VGPR 61% compared to 31% in non-trial patients, and longer PRS, 64 vs 35 months, HR 0.54 (0.36-0.81, p= 0.003) and OS 90 vs 57 months, HR 0.50 (0.33-0.76, p=0.001).

Risk factors at relapse influenced outcomes. Higher ISS was associated with shorter PRS, ISS 2/3 27 vs 50 months for ISS1, HR 2.52(1.74-3.66, p<0.0001) and OS, 46 vs 82 months, HR 2.42(1.67-3.50, p<0.0001). Presence of adverse FISH at relapse (54.1% of 133 patients) also predicted shorter PRS (restricted mean survival time 40 vs 72 months, p<0.001) and OS, 59 vs 97 months for standard FISH, HR 0.46 (0.28-0.73, p=0.0011). Adverse FISH at diagnosis (25.5% of 157) also predicted poorer outcomes.

Conclusions. This real-world series shows that timing of relapse, period of relapse (pre-2008), ISS Stage, adverse FISH, and response to salvage regimen influence survival after relapse from frontline ASCT. Use of novel regimens, particularly PIs as salvage therapy post ASCT is associated with longer OS. Our data confirm the importance of entering patients into clinical trials. Multivariable linear regression analyses will be presented. Lastly, the inferior outcomes for patients with high-risk features highlights the need to develop different treatment strategies in this patient subgroup.