Pooled analysis of long-term data from studies of tandem autologous versus autologous-allogeneic HSCT for newly diagnosed multiple myeloma

Tandem autologous hematopoietic stem cell transplantation (HSCT) (auto-HSCT) has been studied as a method of increasing remission rates and reducing relapse in the upfront treatment of multiple myeloma (MM).^1,2 Autologous HSCT followed by reduced intensity conditioning allogeneic HSCT (auto-allo-HSCT) offers the potential for a long-term graft-versus-myeloma (GVM) effect, but carries a risk of graft-versus-host disease (GvHD) and potentially higher non-relapse mortality (NRM).

Tandem auto-allo-HSCT has been compared with tandem auto-HSCT in multiple prospective studies with conflicting results, leaving clinicians unclear which strategy is preferred for newly diagnosed patients with MM. In 2013, a meta-analysis of published results was performed, which suggested that auto-allo-HSCT was associated with a better chance of a complete response, but with higher NRM and no improvement in PFS or OS compared with tandem auto-HSCT.³

With the passage of time, many of these studies now have long-term follow-up data available, enabling a fresh analysis of pooled individual patient data from these studies. The latest analysis, which included long-term follow-up data from four of the original meta-analyses, was presented in December 2019 at the 61^st American Society of Hematology Meeting & Exposition, Orlando, US, by Luciano Costa from the O’Neal Comprehensive Cancer Center at the University of Alabama, Birmingham, US.⁴

The Multiple Myeloma Hub is covering transplant in myeloma as a monthly theme. Read more here.

Study design

• Studies were selected for inclusion that fulfilled the following criteria:
  • Newly diagnosed patients with MM who had received brief induction therapy
  • Allocation to auto-auto-HSCT or auto-allo-HSCT arms based exclusively on the availability of HLA-matched sibling donors (“biologic randomization”)
  • Conditioning regimen for allo-HSCT met Center for International Blood and Marrow Transplant Research^® (CIBMTR) criteria for reduced intensity
• Investigators from all trials meeting eligibility criteria were invited to submit a limited dataset (age, gender, risk profile, treatment arm, relapse, NRM, PFS, OS)
• Patients were designated high-risk if they had a β2 microglobulin level at diagnosis of ≥ 4 mg/L and/or a deletion of chromosome 13 by metaphase karyotyping
• Outcomes analyzed: PFS, OS, NRM, risk of relapse
• All analyses were by intention-to-treat (ITT) analysis
• NRM and relapse/progression were analyzed as competing risks
• Post-relapse survival (PRS) was defined as the time from relapse/progression until death, censored at last follow-up if alive

Patient characteristics

• Data from four studies were included: BMT CTN 0102, NMAM2000, PETHEMA/GEM2000, and NCT00415987
• 1,338 patients were included in the analysis (Table 1)
  • 899 patients underwent auto-auto-HSCT
  • 439 patients underwent auto-allo-HSCT
• Median follow-up of survivors: 118.5 months
• Patients receiving auto-allo-HSCT were slightly younger than patients receiving auto-auto-HSCT
• The median follow-up of survivors was longer in the auto-allo-HSCT group than in the auto-auto-HSCT group

 Table 1. Patient characteristics

Allo, allogeneic; Auto, autologous; HSCT, hematopoietic stem cell transplantation
	Auto-auto-HSCT (n= 899)	Auto-allo-HSCT (n= 439)
Median follow-up of survivors, months	112.2	122.3
Median age, years	56	53.4
Age < 50 years	238 (26.4%)	150 (34.2%)
Male	527 (58.6%)	249 (56.7%)
High-risk	125 (13.9%)	89 (20.3%)

Results

• PFS and OS were significantly improved after auto-allo-HSCT compared with auto-auto-HSCT (Table 2)

Table 2. OS, PFS, and NRM in newly diagnosed patients with MM receiving either auto-auto-HSCT or auto-allo-HSCT in four studies

Allo, allogeneic; Auto, autologous; CI, confidence interval; HSCT, hematopoietic stem cell transplantation; NRM, non-relapse survival; PFS, progression-free survival; OS, overall survival
	Auto-auto-HSCT (n= 899)	Auto-allo-HSCT (n= 439)	HR (95% CI) p value
OS
Median OS, months (95% CI)	78 (71.5–84.5)	98.3 (81.8–114.7)	HR= 0.84 (0.73–0.97) p= 0.02
5-year OS, % (95% CI)	59.8 (56.6–63)	62.3 (57.8–66.8)	p= 0.37
10-year OS, % (95% CI)	36.4 (32.9–40)	44.1 (39.2–49)	p= 0.01
PFS
Median PFS, months (95% CI)	26.4 (23.8–28.9)	24.4 (18.8–30)	HR= 0.85 (0.75–0.95) p= 0.004
5-year PFS, % (95% CI)	23.4 (20.7–26.1)	30.1 (25.8–34.4)	p= 0.01
10-year PFS, % (95% CI)	14.4 (11.8–16.9)	18.7 (15.0–22.4)	p= 0.06
Relapse
5-year relapse, % (95% CI)	69.7 (66.8–72.6)	52.4 (47.9–56.9)
10-year relapse, % (95% CI)	77.2 (74.5–79.9)	61.6 (56.9–66.3)
NRM
5-year NRM, % (95% CI)	6.9 (5.3–8.5)	17.4 (13.9–20.9)
10-year NRM, % (95% CI)	8.3 (6.5–10)	19.7 (16—23.4)

• For the 214 high-risk patients (125 auto-auto-HSCT, 89 auto-allo-HSCT), 5-year and 10-year PFS rates were significantly higher with auto-allo-HSCT than with auto-auto-HSCT, but there was no difference in OS
• Median PRS was 41.5 months (95% CI, 36.5–46.4) in the auto-auto-HSCT arm and 62.3 months (95% CI, 47.7–76.9) in the auto-allo-HSCT arm (HR= 0.70; 95% CI, 0.58–0.84; p< 0.001)

Conclusions

The authors of this study have concluded that, compared with auto-auto-HSCT, auto-allo-HSCT is associated with the best long-term outcomes, although the benefits are only evident with long-term follow-up. Auto-allo-HSCT was associated with a lower risk of relapse, longer OS and PFS, and a higher risk of NRM compared with auto-auto-HSCT. Robust improvements in post-relapse survival were attributed to potential cooperation between GVM and novel MM agents. Nevertheless, despite the advantages of auto-allo-HSCT over auto-auto-HSCT demonstrated in this study, the authors suggest that the improved long-term survival benefit should be balanced against the late morbidity such as chronic GvHD.