Expert Commentary | The impact of extramedullary disease on transplantation outcomes in multiple myeloma

This month, the Multiple Myeloma (MM) Hub are focusing on the topic of stem cell transplantation. This article is an expert commentary written by Nico Gagelmann, University Medical Center Hamburg-Eppendorf, who, alongside his colleagues, recently published two papers on the impact of extramedullary disease (EMD) on transplant outcomes in patients with newly diagnosed multiple myeloma (NDMM). The first was published in May 2018 in Haematologica and was a study on behalf of the Chronic Malignancies Working Party (CMWP) of the European Society for Blood and Marrow Transplantation (EBMT). The second was published in Biology of Blood and Marrow Transplantation in July 2019 and was also a study by the CMWP of the EBMT, investigating the impact of EMD on the outcomes of patients with high-risk cytogenetics undergoing tandem autologous stem cell transplantation (ASCT).

How does EMD affect patients with NDMM undergoing ASCT?

MM is characterized by a clonal proliferation of malignant plasma cells with a strong dependence on the bone marrow (BM) microenvironment.^1,2 However, in some MM patients, myeloma cells escape the BM, resulting in EMD, which can be further characterized by two different types of involvement²:

1. Paraskeletal (PS), consisting of masses that arise from bone lesions
2. Extramedullary organ (EM), resulting from hematogenous spread into different organs, skin and lymph nodes

The evidence of the effect of EMD at diagnosis has been limited over the years due to small patient numbers, heterogenous patient or intervention selection, and outcome evaluations, mainly in the relapse setting.²

The EBMT therefore investigated EMD in patients with NDMM and its impact on outcome following first-line ASCT.³ The study identified 3,744 patients with available information on the presence or absence of EMD who received upfront transplantation between 2005 and 2014. The overall incidence of EMD was 18.2% (n= 682) and increased each year, from 6.5% (2005) to 23.7% (2014). PS involvement was found in 543 (14.5%) and EM involvement in 139 (3.7%). Involved organs were: kidneys (27%), skin (23%), lymph nodes (17%), central nervous system (10%), lung and respiratory tract (7%), gastrointestinal tract and liver (6%), pleura and heart (5%), and spleen, ovaries and testes (5%). More patients with extramedullary organ involvement had multiple involved sites (≥ two).

In multivariate outcome analysis, organ involvement with ≥ two sites was associated with worse three-year progression-free survival (PFS), with an hazard ratio (HR) of 3.40. Interestingly, there was no difference in PFS in organ involvement with one site compared to NDMM without EMD (HR= 1.03). Comparison of PFS between PS involvement and NDMM without EMD showed no difference for one site with a HR of 1.02, and worse PFS with a HR of 2.46 for ≥ two sites.

In the overall survival (OS) analysis, organ involvement was associated with worse outcome. The difference between patients with one site of PS involvement and those without EMD was less pronounced, with a HR of 1.33 (95% CI, 0.98–1.83; p= 0.07), while ≥ two sites resulted in a similar outcome, but are difficult to interpret owing to the small number of patients.

These results were supported by a recent study from the Balkan Myeloma Study Group and Barcelona University reflecting that EMD per se may not confer worse outcome after upfront transplant⁴, while patients with organ involvement surely define a very high-risk population.

Is tandem ASCT in patients with NDMM and EMD recommended?

Raising dose intensity through use of tandem transplantation with high-dose melphalan (HDM) for NDMM was introduced before the era of novel agents. Several phase III studies have demonstrated a PFS benefit with HDM and some benefit in OS, especially in patients achieving less than complete response (CR) after the first transplant.^5,6 Two contemporary phase III studies evaluating the role of tandem transplantation with HDM following induction therapy reached differing conclusions. Uncertainty remains regarding which groups of patients benefit from tandem transplantation in an era of highly active induction regimens.^7,8 A pooled analysis of phase III studies using induction with bortezomib, thalidomide, and dexamethasone (VTd) or doxorubicin, bortezomib, and dexamethasone (PAD) with prespecified randomization to single or double transplantation reported improved outcome with double transplantation.⁹

With regards to specific NDMM sub-populations, only limited data are available regarding the optimal transplant strategy in NDMM patients with EMD and high-risk cytogenetics. Current consensus defines t(4;14), t(14;16), t(14;20), del(17p), del(1p), and gain(1q) and any non-hyperdiploid karyotype as high-risk cytogenetics in patients with MM, resulting in poor outcome.¹⁰ However, currently existing and newly proposed cytogenetic risk classification often includes a mix of therapies,^10,11 while recent studies have highlighted that the outcome of patients with different cytogenetic abnormalities is affected by treatment (e.g. transplant versus no transplant).¹² Subsequently, the evolving treatment landscape in MM aims to identify specific subgroups that may benefit the most by receiving specific therapy.

A recent analysis from the EBMT focused on patients with NDMM with EMD specifically, with available cytogenetic information from BM samples, which could be obtained from 488 patients who underwent upfront hematopoietic stem cell transplantation.¹³ The study demonstrated that 40% of patients with NDMM with EMD had high-risk cytogenetics, which is significantly more frequent than usually seen in patients with NDMM without EMD. At least one high-risk cytogenetic abnormality was present in 202 patients (41%), with del(17p) and t(4;14) being the most frequent in these patients (40% and 45%, respectively). Of these 202 patients, a single high-risk cytogenetic abnormality was found in 46% of all, while the remaining 54% had more than one high-risk abnormality.¹³

Four-year OS appeared to be influenced by transplant in univariate analysis of these data, resulting in OS rates of 70% (63–76%) for single ASCT versus 83% (74–92%) for tandem ASCT and 88% (74–100%) for autologous-allogeneic stem cell transplant (auto-allo SCT) (p= 0.06). Four-year PFS was 43% (37–49%) for single ASCT versus 52% (40–64%) for tandem ASCT and 58% (35–81%) for auto-allo SCT at four years (p= 0.30). The cumulative incidence of non-relapse mortality was 2% for single ASCT, 1% for tandem ASCT, and 10% for auto-allo SCT (p= 0.09). The corresponding cumulative incidences of relapse were 54%, 47%, and 30% (p= 0.29).¹³ 

After single ASCT, outcome was significantly different for both OS and PFS in patients at different risk according to cytogenetics, resulting in 78% (71–85%) and 48% (40–56%) for standard-risk versus 41% (30–52%) and 22% (11–33%) for high-risk cytogenetics, respectively (p< 0.001). In contrast, tandem ASCT did not show differences between standard- and high-risk cytogenetics in OS or PFS being 82% (70–95%) and 56% (39–73%) for standard-risk versus 84% (71–97%) and 45% (26–64%) for high-risk (p= 0.99 and p= 0.24).¹³

In a multivariate analysis, high-risk cytogenetics were associated with worse survival (HR= 2.00; 95% CI, 1.28–3.15; p= 0.003), while tandem ASCT significantly improved outcome versus single ASCT (HR= 0.46; p= 0.02). Auto-allo SCT did not significantly differ in outcome but appeared to improve survival, while results were limited due to small population (HR= 0.31; p= 0.06).¹³

Allo-SCT has been proposed as a treatment of younger and fit patients at high risk, but data on cytogenetic abnormalities are scarce. The EBMT study provides retrospective data, and thus there is some limited evidence that carefully selected patients with NDMM and EMD may benefit from an upfront auto-allo SCT approach, especially those carrying high-risk cytogenetics. These results regarding the role of allograft in high-risk patients are further supported by a trial of 73 patients with NDMM, tandem auto-allo SCT yielded similar five-year outcomes in patients with and without t(4;14) or del(17p).¹⁴ A long-term follow-up (median: 91 months) and cytogenetic subgroup analysis of a prospective trial comparing upfront tandem ASCT and auto-allo SCT in patients with del(13q) showed a median PFS of 22 months compared with 35 months (HR= 0.55; 95% CI, 0.36–0.84; p= 0.003), and similar OS (72 vs 70 months; p= 0.856). In patients harboring both del(13q) and del(17p), outcome was significantly improved after auto-allograft, while this subgroup analysis was limited by low number of patients (auto-allo SCT: 19, tandem ASCT: 6).¹⁵ Overall, however, more investigations of upfront tandem auto-allo SCT in patients with NDMM, specifically with EMD, are warranted.

Conclusion

Taken together, outcome in NDMM with EMD is influenced, first, by different sites and number of involvements. Further research should focus on accurate diagnosis using advanced imaging techniques (PET-CT, MRI) as well as novel treatment approaches. Notably, promising new drugs such as daratumumab or pomalidomide failed to show efficacy in relapsed/refractory EMD patients while anti-B-cell maturation antigen (BCMA) chimeric antigen receptor (CAR) T-cell therapies showed astonishing response rates and outcome data in highly pretreated patients with EMD. Secondly, as two other upcoming abstracts from the Annual Meeting of the American Society of Hematology will highlight (Schönland et al. #324 and Costa et al. #259), tandem ASCT shows improved outcome (at least in PFS) in patients with NDMM, while tandem ASCT-allo-SCT transplant provides long-term benefit for high-risk and fit patients. However, data from the EBMT on improved outcome after tandem ASCT in EMD with high-risk cytogenetics show that careful identification of specific subgroups is crucial. Myeloma, and with it EMD, is and will be a heterogeneous disease from the time of diagnosis.