Little progress has been made in the management of the 15% of patients presenting with GEP-70-defined high risk MM. Successive TT protocols (TT3, TT5) were reviewed and clinical endpoints (CR, CR duration, PFS, OS) compared and prognostic models developed for this high-risk entity. Altogether 127 patients were identified (TT3a, 40; TT3b, 37; TT5, 50), whose baseline characteristics were similar with regard to age, B2M, CRP, creatinine, hemoglobin, platelet count, LDH, metaphase-based cytogenetic abnormalities and the recently introduced GEP-5 risk designation (Table 1 

Table 1

Patient characteristics, TT3 vs. TT5

FactorTT3TT5P-value
Age >= 65 yr 21/77 (27%) 14/50 (28%) 0.929 
Albumin < 3.5 g/dL 44/77 (57%) 25/49 (51%) 0.501 
B2M >= 3.5 mg/L 57/77 (74%) 40/49 (82%) 0.323 
B2M > 5.5 mg/L 41/77 (53%) 22/49 (45%) 0.361 
CRP >= 8 mg/L 35/77 (45%) 18/49 (37%) 0.334 
Creatinine >= 1.5 mg/dL 18/77 (23%) 11/49 (22%) 0.904 
Hb < 10 g/dL 41/77 (53%) 28/49 (57%) 0.668 
LDH >= 190 U/L 37/77 (48%) 17/49 (35%) 0.140 
Platelet Count < 150 x 10^9/L 24/77 (31%) 17/49 (35%) 0.681 
Cytogenetic abnormalities 57/76 (75%) 33/49 (67%) 0.352 
GEP-5 High-Risk 58/77 (75%) 34/50 (68%) 0.367 
FactorTT3TT5P-value
Age >= 65 yr 21/77 (27%) 14/50 (28%) 0.929 
Albumin < 3.5 g/dL 44/77 (57%) 25/49 (51%) 0.501 
B2M >= 3.5 mg/L 57/77 (74%) 40/49 (82%) 0.323 
B2M > 5.5 mg/L 41/77 (53%) 22/49 (45%) 0.361 
CRP >= 8 mg/L 35/77 (45%) 18/49 (37%) 0.334 
Creatinine >= 1.5 mg/dL 18/77 (23%) 11/49 (22%) 0.904 
Hb < 10 g/dL 41/77 (53%) 28/49 (57%) 0.668 
LDH >= 190 U/L 37/77 (48%) 17/49 (35%) 0.140 
Platelet Count < 150 x 10^9/L 24/77 (31%) 17/49 (35%) 0.681 
Cytogenetic abnormalities 57/76 (75%) 33/49 (67%) 0.352 
GEP-5 High-Risk 58/77 (75%) 34/50 (68%) 0.367 

n/N(%): n- Number with Factor, N-Number with Valid Data for Factor

N/A:All Positive or All Negative for Factor, P-value not available

* Fisher Exact Test, otherwise Chi-Square Test

View Large
). Compared to TT3a/b combined, TT5 resulted in improved OS (2-yr estimate 92% v 68%, p=0.02) with no difference noted for PFS and CR duration. GEP-5-defined high-risk MM fared significantly worse for OS (p=0.02) and PFS (p=0.04), without difference in CR duration (p=0.4). Univariate Cox regression analysis was performed on standard prognostic factors as well as time-dependent indicators for achieving CR and disease progression; the interaction term for achieving CR and disease progression (i.e. relapse from CR) was also considered for the multivariate model (Table 2). A stepwise selection algorithm identified age >=65yr and disease progression as independent adverse variables, while TT5 had a HR=0.31 (p=0.002). The results from this model highlight the importance of improving the rate of progression in high-risk MM. The importance of sustained CR or progression-free survival for at least 2 years is demonstrated in Figure 1 . Patients relapsing within 2 years from CR or lower response level survived a subsequent median of less than 1 year.
Table 2

Multivariate Cox Regression Analysis

Overall Survival
Variablen/N (%)HR (95% CI)P-value
Multivariate Model TT5 48/124 (39%) 0.29 (0.14, 0.60) <.001 
Disease Progression ***  13.13 (7.47, 23.10) <.001 
Age >= 65 yr 34/124 (27%) 2.44 (1.38, 4.31) 0.002 
Multivariate Model with Additional Term for Achieving CR TT5 50/127 (39%) 0.30 (0.14, 0.64) 0.002 
Achieved CR ***  0.63 (0.35, 1.11) 0.110 
Disease Progression ***  11.99 (6.85, 21.00) <.001 
Age >= 65 yr 35/127 (28%) 2.50 (1.41, 4.43) 0.002 
Overall Survival
Variablen/N (%)HR (95% CI)P-value
Multivariate Model TT5 48/124 (39%) 0.29 (0.14, 0.60) <.001 
Disease Progression ***  13.13 (7.47, 23.10) <.001 
Age >= 65 yr 34/124 (27%) 2.44 (1.38, 4.31) 0.002 
Multivariate Model with Additional Term for Achieving CR TT5 50/127 (39%) 0.30 (0.14, 0.64) 0.002 
Achieved CR ***  0.63 (0.35, 1.11) 0.110 
Disease Progression ***  11.99 (6.85, 21.00) <.001 
Age >= 65 yr 35/127 (28%) 2.50 (1.41, 4.43) 0.002 

HR-Hazard Ratio, 95% CI- 95% Confidence Interval, P-value from Wald Chi-Square Test in Cox Regression. All univariate pvalues reported regardless of significance.

Multivariate model uses stepwise selection with entry level 0.1 and variable remains if meets the 0.05 level.

A multivariate p-value greater than 0.05 indicates variable forced into model with significant variables chosen using stepwise selection.

*** Indicates Time-dependent variable.

View Large

Figure 1
Figure 1. Overall Survival by Response Status at 2 year Landmark
View largeDownload slide

Overall Survival by Response Status at 2 year Landmark

Figure 1
Figure 1. Overall Survival by Response Status at 2 year Landmark
View largeDownload slide

Overall Survival by Response Status at 2 year Landmark

Close modal

Thus, GEP-70-defined high-risk MM has not yet benefited from therapeutic advances observed in low-risk disease. The superior OS in TT5 may be due to the availability of additional new agents, pomalidomide and carfilzomib, which will be incorporated into the front line management of such patients. A plateau-like shape of the PFS curve emerged at 4 years supporting previous observations in TT2 that GEP-defined high-risk MM can be cured in a fraction of patients. The GEP-5 model points to the detrimental role of a limited number of genes which, when targeted, may offer means of controlling this disease entity. We are also applying exomics to identify, in high-risk MM, early on actionable mutations for personalized therapy of such patients.

Disclosures:

van Rhee:Jansen & Jansen: Research Funding. Barlogie:Celgene: Consultancy, Honoraria, Research Funding; Myeloma Health, LLC: Patents & Royalties.

Topics:
gene expression profiling, multiple myeloma, disease progression, c-reactive protein, creatinine, cyclic amp receptor protein, epley maneuver, albumins, carfilzomib, hemoglobin

Author notes

*

Asterisk with author names denotes non-ASH members.

This icon denotes a clinically relevant abstract

© 2013 by The American Society of Hematology
2013
Sign in via your Institution