Using a novel model-based projection method, researchers have announced that long-term survival rates for young multiple myeloma patients are not only higher than ever before, but also higher than current estimates suggest.

Patients under age 45 who are diagnosed in 2006-2010 can expect a 5-year relative survival rate of 68 percent and a 10-year relative survival rate of 55 percent. Relative survival reflects the survival of multiple myeloma patients compared to the general population. These values range from seven to 20 percentage points higher than standard, currently-used estimates.

Using data from the 1973-2005 Surveillance, Epidemiology, and End Results (SEER) cancer database, researchers evaluated the new model-based prediction method against the traditional methods of cohort and period analysis, two alternatives for analyzing survival outcomes. The novel method, because it is forward-looking, can provide up-to-date estimates without waiting the five or more years for new data to enter the SEER registry. By avoiding such delay, it may better reflect current survival rates, especially as new treatments continually become available.

Application of the modeling approach consistently yielded 5- and 10-year relative survival estimates that were higher than the most up-to-date predictions obtained by period or cohort analysis. In addition, when applied to past data, the novel method’s estimates more closely comported with the documented relative survival rates in the database.

Overall, the greatest differences in survival estimates between the three methods occurred in young patients below 45 years of age. According to the novel model, more than two-thirds of these patients can expect to be alive in five years, whereas period and cohort analysis only predict 61 percent and 53 percent, respectively. At 10 years, the novel method predicts that more than half of patients will still be alive.

In older age groups, especially those above 65 years of age, long-term survival estimates remain rather poor, and differences between the three analysis methods were generally minor.

Although this new model appears to reflect up-to-date survival estimates better than traditional methods, researchers caution that because the SEER registry does not include medication information, they cannot directly link survival estimates to different treatments. In addition, whether or not these projections hold true will require waiting until at least the year 2018. Nevertheless, the study authors predict that these more optimistic survival estimates for young individuals are indeed accurate and encourage physicians to share this information with their patients.

For more information, please see the full article in the February 2009 issue of Haematologica ^[1] journal.