Background: New therapeutic options have led to substantial increases in survival expectations of patients with multiple myeloma (MM) in recent years. In the past, the impact of these innovations on long-term survival of MM patients on the population level has been disclosed only with substantial delay, partly due to delayed availability of data from population-based cancer registries.

Methods: Using data from the 1973–2005 database of the Surveillance, Epidemiology, and End Results (SEER) Program, we employed a novel model-based projection method to project 5-and 10-year relative survival expectations of MM patients in the United States diagnosed in 2006 to 2010, after conducting an empirical evaluation of the method using data from patients diagnosed between 1996 and 2000 (the last cohort for which 5-year follow up is available through SEER) and earlier years.

Results: Preliminary empirical evaluation of the method using historical data indicates good performance in projection of age specific and overall 5- and 10-year relative survival. Model based projection of survival produced survival estimates that matched observed survival estimates more closely than cohort analysis or period analysis, the standard methods of survival analysis, in 21 out of 24 cases. The average difference from observed was −1.27 percentage units for model based projection, −3.71 percentage units for period analysis, and −4.23 percentage units for cohort analysis. Five-year relative survival projections for patients diagnosed in 2006–2010 at ages 15–49 and 50–59 are 61.4% and 52.0% (see table), respectively, and exceed the most up-to-date estimates obtained from traditional cohort analysis by 9.5 and 6.9 percent units, respectively. Ten-year relative survival projections for patients in these age groups are 47.1% and 30.4%, exceeding the most up-to-date estimates from traditional cohort analysis by 13.5 and 9.6 percent units, respectively. By contrast, survival projections remain much lower and hardly exceed estimates from traditional survival analysis for older patients.

Conclusions: The use of model based projection results in more up-to-date estimates of survival than standard methods of survival analysis. Patients diagnosed with multiple myeloma in 2006–2010 have much higher long-term survival expectations than suggested by previously available survival statistics.

Table: Most up-to-date estimates of 5-year and 10-year relative survival of patients diagnosed with multiple myeloma obtained by different methods.

AgeCohort estimates*Period estimates for 2001–2005 periodProjections for patients diagnosed in 2006–2010
PESEPESEPESE
* The most up-to-date cohort estimates of 5-year survival pertain to patients diagnosed in 1996–2000, and the most up-to-date cohort estimates of 10-year survival pertain to patients diagnosed in 1991–1995. 
5-year relative survival 
15+ 32.1 0.7 34.2 0.7 36.1 1.0 
15–49 51.9 2.3 56.4 2.2 61.4 3.1 
50–59 45.1 1.7 48.0 1.6 52.0 2.3 
60–69 33.8 1.4 36.5 1.4 36.7 2.0 
70–79 27.4 1.2 26.7 1.2 27.9 1.7 
80+ 13.2 1.4 14.1 1.4 16.4 1.9 
10-year relative survival 
15+ 14.1 0.6 16.6 0.6 18.0 0.8 
15–49 33.6 2.2 40.3 2.4 47.1 3.2 
50–59 20.8 1.6 27.0 1.7 30.4 2.3 
60–69 12.4 1.0 15.0 1.2 15.4 1.4 
70–79 8.8 1.0 8.9 0.9 9.9 1.1 
80+ 6.4 1.6 5.9 1.4 6.6 1.4 
AgeCohort estimates*Period estimates for 2001–2005 periodProjections for patients diagnosed in 2006–2010
PESEPESEPESE
* The most up-to-date cohort estimates of 5-year survival pertain to patients diagnosed in 1996–2000, and the most up-to-date cohort estimates of 10-year survival pertain to patients diagnosed in 1991–1995. 
5-year relative survival 
15+ 32.1 0.7 34.2 0.7 36.1 1.0 
15–49 51.9 2.3 56.4 2.2 61.4 3.1 
50–59 45.1 1.7 48.0 1.6 52.0 2.3 
60–69 33.8 1.4 36.5 1.4 36.7 2.0 
70–79 27.4 1.2 26.7 1.2 27.9 1.7 
80+ 13.2 1.4 14.1 1.4 16.4 1.9 
10-year relative survival 
15+ 14.1 0.6 16.6 0.6 18.0 0.8 
15–49 33.6 2.2 40.3 2.4 47.1 3.2 
50–59 20.8 1.6 27.0 1.7 30.4 2.3 
60–69 12.4 1.0 15.0 1.2 15.4 1.4 
70–79 8.8 1.0 8.9 0.9 9.9 1.1 
80+ 6.4 1.6 5.9 1.4 6.6 1.4 

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution