Despite the use of novel drugs and autologous stem cell transplantation (SCT), multiple myeloma (MM) is still considered to be an incurable disease. Both autologous and allogeneic SCT (allo-SCT) improve long-term survival. Allo-SCT can result in long-term disease control and remains a potentially curative therapeutic modality, related in part to the immune anti-myeloma effect of the allograft, but it is hampered by high transplant-related toxicities. Reduced intensity conditioning (RIC) allo-SCT has been developed in order to reduce transplant-related mortality while maintaining a graft-versus-myeloma (GVM) effect.
Few studies have reported long-term outcomes in MM after allo-SCT, with a median follow-up typically ranging from 2 to 5 years. Recently, Sahebi et al (2013) reported an overall survival (OS) of 60%, progression-free survival (PFS) of 31% and relapse incidence of 59% at 7 years in 60 MM patients undergoing allo-SCT after RIC, with a median follow-up of 9·8 years. Very late relapses occurred in 10% of patients, even as late as 11·5 years after transplantation (Sahebi et al, 2013).
We report on the long-term follow-up (median 6·82 years) of 42 consecutive MM patients who underwent myeloablative conditioned (MAC) or RIC allo-SCT at a single centre between 1988 and 2011. Baseline characteristics of the patients are detailed in Table 1. Most of the patients had advanced stage and standard risk cytogenetic abnormalities. Conventional chemotherapy was given as induction in all but two. At the time of transplantation, 28 patients had chemo-sensitive disease in complete (CR, n = 4) or partial remission (PR, n = 24) whereas 14 were in relapse or progression (PD). Using bone marrow (n = 16) or peripheral stem cells (n = 26) from a familial (n = 37) or an alternative (n = 5) matched donor, the allografting procedure was either scheduled as part of upfront treatment (n = 21), as a planned tandem auto-allo-transplant (n = 7), or as salvage therapy for relapsed disease (n = 21) or after failure of auto-SCT (n = 7). Thirteen patients received donor lymphocyte infusions (DLI).
|Variable||N (%), median (range)|
|Age at transplant (years)||51 (36–67)|
|< 50||3 (7)|
|Disease stage at diagnosis|
|Stage I||6 (14)|
|Stage II||12 (29)|
|Stage III||24 (57)|
|Cytogenetic risk stratification|
|Standard riska||20 (48)|
|High riskb||8 (19)|
|Time from diagnosis to transplant|
|<1 year||21 (50)|
|>1 year||21 (50)|
|First line||21 (50)|
|Salvage therapy||21 (50)|
|Conventional chemotherapy||32 (76)|
|Novel agents||8 (19)|
|Disease status prior allo-SCT|
|Complete remission||4 (10)|
|Partial remission||24 (57)|
|Stable/progressive disease||14 (33)|
|Stem cell source|
|Bone marrow||16 (38)|
|Peripheral blood stem cell||26 (62)|
|Matched unrelated donor||5|
|Gender match: donor recipient|
|Female donor male recipient||8 (19)|
|Myeloablative conditioning||25 (60)|
|Reduced intensity conditioning||17 (40)|
Outcomes are listed in Table 2. Acute graft-versus-host disease (GVHD) grade II-IV occurred in 18 patients, and 3 had grade III-IV. Seven patients died early and could not be evaluated for chronic GVHD (cGVHD). Extensive cGVHD was observed in 7 of 35 patients (20%); 8 displayed cGVHD (23%) and 20 had no cGVHD (57%). Best responses achieved post-transplant were CR (50%) and PR (38%). Six patients achieved CR after an allograft performed in PD. Three out of 13 patients achieved a PR after DLI.
|Outcome||N (%), median (range)||SE||95% CI|
|Best response to transplant|
|Cause of death|
|Median survival (days)|
|All patients||2490 (209–9642)|
|At 5 years||50%||0·07||0·25–0·56|
|At 10 years||45%||0·07||0·29–0·60|
|At 20 years||7%||0·10||0·06–0·45|
|At 5 years||50%||0·07||0·32–0·62|
|At 10 years||43%||0·07||0·28–0·59|
|At 20 years||7%||0·06||0·00–0·19|
|Cumulative incidence of NRM|
|1 year NRM||24%||0·06||0·13–0·41|
|2 years NRM||26%||0·06||0·15–0·44|
|5 years NRM||29%||0·07||0·18–0·47|
|20 years NRM||29%||0·07||0·18–0·47|
|Cumulative incidence of RM|
|1 year RM||25%||0·07||0·14–0·44|
|2 years RM||40%||0·08||0·26–0·60|
|5 years RM||69%||0·08||0·55–0·88|
|20 years RM||86%||0·07||0·72–1·00|
At the time of analysis, 16 patients are alive, 7 of them in CR. OS was 50%, 45% and 26%, and PFS 50%, 43% and 7% at 5, 10 and 20 years, respectively. There were no significant differences in outcome between RIC and MAC, or between allografts performed in first-line or as salvage therapy, early (<1 year) or late (>1 year) after diagnosis. Chronic GVHD and achievement of CR after SCT were not significantly associated with better OS or PFS.
At the time of analysis, 26 patients have progressed/relapsed. Twenty-six have died, 13 from disease progression or relapse and 13 from transplant-related complications. Non-relapse mortality (NRM) was 24%, 29% and 29% at 1, 5 and 20 years, respectively. The 1-year incidence of NRM was not significantly different between MAC and RIC.
Cumulative incidence of progression/relapse was 25%, 69% and 86%, at 1, 5 and 20 years, respectively, without significant difference between RIC and MAC, or between an allograft performed first-line or as salvage therapy.
Six patients suffered a late relapse, which occurred between 6 and 10·5 years after the procedure. None of these six patients had high-risk features at diagnosis. Only two were transplanted within 1 year from diagnosis and four received a MAC regimen. All experienced an acute GVHD grade I-II. None of them displayed cGVHD nor received consolidation or maintenance therapy after the procedure. One died of refractory MM 12·5 years after transplantation, and two died of metastatic cancers (angiosarcoma and pancreatic adenocarcinoma) at 10 and 15 years after allo-SCT, respectively.
Two further patients experienced malignancies. One patient was diagnosed with prostate carcinoma while under lenalidomide maintenance therapy. The other patient developed an acute myeloid leukaemia in donor cells, 4 years after transplantation, and has been in CR for both diseases for 9 years (Havelange et al, 2006).
Based on the long-term follow-up of this cohort, we confirm the previous observation that more than 10% of myeloma patients experienced late progression/relapse after allo-SCT (Sahebi et al, 2013). Neither cGVHD nor achievement of CR after allo-SCT were significantly associated with improved OS or longer PFS in our series.
In agreement with previous studies emphasizing that toxicity remains a major drawback of the procedure, we observed a 29% incidence of toxic death that was mainly related to organ failure or infection. RIC has been proposed to reduce transplant-related mortality while maintaining a GVM effect (Bruno et al, 2007). However, RIC is associated with a higher risk of relapse (Crawley et al, 2007). Adding novel drugs as part of conditioning and maintenance treatment might also prevent relapse. The use of bortezomib with conditioning might limit the risk of relapse after RIC (Caballero-Velazquez et al, 2013). As it selectively eliminates alloreactive T-cells, it could also be of value in preventing GVHD (Koreth et al, 2009) and avoiding toxicities. Maintenance with lenalidomide reduces relapses, as it potentiates Natural Killer cell activity (Danylesko et al, 2012), but exposes patients to more GVHD (Kneppers et al, 2011). Low-dose lenalidomide (5 mg/day) might be an alternative option (Wolschke et al, 2013). Future studies should explore the integration of novel agents and cellular therapies with RIC in an attempt to reduce the incidence of disease recurrence.
Twenty patients were long-term survivors (more than 3000 days after diagnosis), possibly because of the use of novel agents at relapse and the achievement of PR after DLI in 3 out of 13 cases. In addition, at the date of final analysis, 16 patients are alive, 7 of whom are in CR. The present data underscores that long-term disease control can be achieved in a subset of MM patients undergoing allo-SCT, even when performed in heavily pre-treated, advanced stage, refractory MM.